JP6963819B2 - How the device operates, the device and the system for automatically controlling inhaled oxygen distribution - Google Patents

Description

The present invention generally relates to methods, devices and systems for automatically controlling inhaled oxygen distribution. For example, it relates to methods, devices and systems for automatically controlling the inhaled oxygen concentration so as to maintain the oxygen concentration within the target range.

Oxygen supplementation therapy can be used for a variety of purposes in both chronic and acute patient therapy. For example, it plays a central role in the treatment of newborns with respiratory illness. Studies of preterm infants show a link between chronic hypoxia and increased mortality. Excessive oxygen distribution was also observed to be counterproductive, especially in premature infant retinopathy. Therefore, in order to prevent excess oxygen, it is necessary to continuously adjust the proportion of inhaled oxygen (FiO _{2 ) to maintain the oxygen concentration (SpO 2) within an acceptable range.} The response of SpO _{2 to} changes in FiO ₂ is called the system gain, where the system is the patient.

Currently, it is largely in the hands of bedside healthcare professionals that determines the balance of oxygen distribution for preterm infants. The person manually adjusts _{FiO 2} to maintain the _{oxygen concentration within the SpO 2 target range.} Unfortunately, _{this manual control of FiO 2} is inaccurate, and infants will spend quite a long time on _{SpO 2 outside the target range.}

Automatic regulation of FiO ₂ allows for longer stays in the target range than manual control and can significantly reduce the rate of iatrogenic hyperoxia and severe hypoxia. However, there are important challenges in applying automatic oxygen distribution to preterm infants with pulmonary dysfunction. The first task is _{to improve the effect within the SpO 2} target range and to avoid the time and onset of hypoxia and hyperoxygenosis. The second challenge of automatic control of oxygen distribution is that the major determinants of oxygen supply are mixed by preterm infants and contribute to fundamentally different responses to changes _{in FiO 2 and are therefore given.} An automated controller with a uniform and invariant response to SpO ₂ perturbations is not possible to meet the needs of every individual. The third issue is that the system gain changes over time.

It is hoped that one or more shortcomings or limitations related to the prior art will be resolved or ameliorated, or at least a useful alternative will be provided.

Any prior document (or information derived from it) or a reference to a well-known technique herein refers to a prior art document (or information derived from it) or a well-known technique in the art to which this specification relates. Should not be admitted or proposed to form part of the common general knowledge in.

According to one aspect of the present invention, a method for automatically controlling inhaled oxygen distribution is provided. The method is
The process of receiving signals representing multiple input oxygen concentration (SpO _{2) values for the patient, and}
A process of generating a control value based on an input SpO ₂ value and a target SpO _{2 value, and}
A step of generating an output inspired oxygen concentration (FiO ₂ ) value based on a controlled value and a reference inspired oxygen concentration (rFiO _2).
Have,
The control value is
Input SpO ₂ values, target SpO ₂ values, and immediate control values generated based on the immediate gain factor,
Input SpO ₂ value, target SpO ₂ value, and cumulative control value generated based on the cumulative gain coefficient,
Input SpO ₂ values, target SpO ₂ values, and predicted control values generated based on the predicted gain coefficient,
Have,
The immediate control value is determined based on _{the rFiO 2 value.}
Non-linear weighting compensation is applied to the cumulative control value based on a predetermined non-linear relationship between the partial pressure of arterial oxygen (PaO ₂ ) and SpO _2.

In addition to the present invention, an apparatus for automatically controlling inhaled oxygen distribution is provided according to an embodiment. The device is
An input unit that receives signals representing multiple input oxygen saturation (SpO _{2) values for the patient, and}
A memory for recording the received input SpO _{2 value, and}
A controller for determining the output oxygen concentration (FiO ₂ ) value based on the input SpO _{2 value,}
An output unit for outputting the determined output FiO _{2 values,}
Have,
The controller is
Generate a control value based on the input SpO ₂ value and the target SpO _{2 value.}
Generate an output inspired oxygen concentration (FiO ₂ ) based on the control value and the reference inspired oxygen concentration (rFiO _{2) value.}
The control value is
Input SpO ₂ values, target SpO ₂ values, and immediate control values generated based on the immediate gain factor,
Input SpO ₂ value, target SpO ₂ value, and cumulative control value generated based on the cumulative gain coefficient,
Input SpO ₂ values, target SpO ₂ values, and predicted control values generated based on the predicted gain coefficient,
Have,
The immediate control value is determined based on _{the rFiO 2 value.}
Non-linear weighting compensation is applied to the cumulative control value based on a predetermined non-linear relationship between the partial pressure of arterial oxygen (PaO ₂ ) and SpO _2.

According to another aspect of the present invention, a system for automatically controlling inhaled oxygen distribution is provided. The system is
One or more oxygen saturation monitoring devices and one or more inhaled oxygen control devices,
Control device and
A network that allows communication between one or more oxygen saturation monitoring devices and control devices, and between one or more inhaled oxygen control devices and control devices.
Have,
The control device is
Through the network, each of one or more oxygen saturation monitoring devices _{receives signals representing multiple input oxygen saturation (SpO 2} ) values for the patient.
Generate a control value based on the input SpO ₂ value and the target SpO _{2 value.}
Generate an output inspired oxygen concentration (FiO ₂ ) value based on the controlled and reference inspired oxygen concentration (rFiO _{2) values.}
Send the _{determined output FiO 2} value to the corresponding inspired oxygen control device over the network.
By controlling the inhaled oxygen distribution,
The control value is
Input SpO ₂ values, target SpO ₂ values, and immediate control values generated based on the immediate gain factor,
Input SpO ₂ value, target SpO ₂ value, and cumulative control value generated based on the cumulative gain coefficient,
Input SpO ₂ values, target SpO ₂ values, and predicted control values generated based on the predicted gain coefficient,
Have,
The immediate control value is determined based on _{the rFiO 2 value.}
Non-linear weighting compensation is applied to the cumulative control value based on a predetermined non-linear relationship between the partial pressure of arterial oxygen (PaO ₂ ) and SpO _2.

According to another aspect of the present invention, a method for automatically controlling inhaled oxygen distribution is provided. The method is
The process of receiving signals representing multiple input oxygen saturation (SpO _{2) values for the patient, and}
A process of generating a control value based on an input SpO ₂ value and a target SpO _{2 value, and}
A step of generating an output inspired oxygen concentration (FiO ₂ ) value based on a controlled value and a reference inspired oxygen concentration (rFiO _{2) value.}
Have,
The control value is
Input SpO ₂ values, target SpO ₂ values, and immediate control values generated based on the immediate gain factor,
Input SpO ₂ value, target SpO ₂ value, and cumulative control value generated based on the cumulative gain coefficient,
Input SpO ₂ values, target SpO ₂ values, and predicted control values generated based on the predicted gain coefficient,
Have.

Some embodiments of the present invention are exemplified below with reference to the accompanying drawings.

FIG. 1 is a block diagram of an inhaled oxygen distribution system. FIG. 2 is a block diagram of the components in the inhaled oxygen distribution system. FIG. 3 is a flowchart showing the process of generating the PID term, that is, the proportional term, the integral term, and the differential term. FIG. 4 is a flowchart showing a process of modifying _Kp based on the performance evaluation result. FIG. 5 is a flowchart showing a process of modifying the value of _{rFiO 2.} FIG. 6 is a flowchart showing the process of determining the perfusion index value. FIG. 7 is a flowchart showing the process _{of the hierarchical SpO 2 verification method.} FIG. 8 is a flowchart showing a process of determining the _{output FiO 2} value based on the effectiveness of _{SpO 2.} FIG. 9 is a flowchart showing a process of switching between the manual mode and the automatic control mode. FIG. 10 is a flowchart showing the process of controlling and resetting the alarm. FIG. 11 is a diagram of a user interface of the oxygen distribution system apparatus. FIG. 12 is a flow chart showing the overall control process of the method for automatically controlling inhaled oxygen distribution. FIG. 13 is a flowchart showing the process of processing the input. FIG. 14 is a flowchart showing the process of reading the input. FIG. 15 is a flowchart showing the process of verifying the input. FIG. 16 is a flowchart showing an input limit verification process. FIG. 17 is a flowchart showing the process of automatic control. FIG. 18 is a flowchart showing a periodic adaptation process. FIG. 19 is a graph showing the relationship between the _{PaO 2} error and the unitary SpO _{2 error.} FIG. 20 is a block diagram of the inhaled oxygen distribution system used in the second exemplary experiment. FIG. 21A is a graph of 2 hour recordings from the same infant during manual and automatic control recorded in the second exemplary experiment. FIG. 21B is a graph of 2 hour recordings from the same infant during manual and automatic control recorded in the second exemplary experiment. FIG. 22 is a graph of a frequency histogram of _{pooled SpO 2} data according to the results of a second exemplary experiment. FIG. 23 is a graph comparing automatic control and best manual control time according to the results of the second exemplary experiment.

The inhaled oxygen distribution system 100 described herein implements a method of automatically controlling inhaled oxygen distribution to a patient (eg, a human infant).

The system (and method) described provides one or more advantages over existing systems and methods. First, the described system _{efficiently targets the desired SpO 2} range and avoids over-inhalation of oxygen. Second, the system described can respond quickly to, for example, changes in V / Q ratios and SpO _{2 deviations due to shunts in the lungs.} Third, the system described compensates for non-linearity within the _{PaO 2-} SpO _{2 relationship (where PaO 2} means partial pressure of oxygen in the artery, for example PaO ₂ is an S-shaped curve. _{For each 1% step change of SpO 2} in the linear portion of, it changes by 1-2 mmHg, but further in the direction of the asymptote by more than 20 mmHg.) Fourth, the system described is: _{Shift of the FiO 2-} SpO ₂ curve (where the right shift is the decreasing ventilation-perfusion (V / Q) ratio) for different individuals to compensate for the individual's variable SpO ₂ response to FiO _{2 regulation.} Can respond differently in response to a mixture of individuals with different (corresponding to) and shunts (percentage or ratio of blood pumped into the human body without oxygenation inside the lungs). Fifth, the system described can adjust its gain based on performance measurements.

As shown in FIG. 1, the system 100 includes a control device 10, an oximeter 20, and a respiratory support device 30.

The control device 11 is configured to automatically control the inhaled oxygen distribution.

The oximeter 20 measures the arterial oxygen saturation (SpO ₂ ) value of the patient 40 and transmits an output signal representing the _{SpO 2} value to the control device 10. _{The SpO 2} value represented by the output signal of the oximeter 20 is also referred _{to as an input SpO 2} value from the viewpoint of the control device 10.

The oximeter 20 has an analog or digital data output.

_{Based on the input SpO 2} value from the oximeter 20, the controller 10 determines the output inspired oxygen concentration (FiO ₂ ) value and outputs a FiO ₂ signal _{representing the determined output FiO 2 value.}

_{The output FiO 2} signal from the control device 10 is transmitted to the respiratory support device 30. The respiratory support device 30 is a system capable of responding to _{FiO 2 inputs.} That is, the respiratory support system can receive and execute the desired value of _{FiO 2.} The respiratory support device 30 may be in the form of an air-oxygen mixer, mechanical ventilator, continuous positive airway pressure (CPAP) driver, or flow generator for high-flow nasal cannula support or low-flow oxygen distribution.

The respiratory support device 30 _{distributes a mixed gas (slightly inhaled oxygen with determined FiO 2} ) to the patient 40. The inhaled oxygen distribution system 100 further includes an auxiliary patient monitoring system 50 and a respiratory circuit monitoring system 60.

The auxiliary patient monitoring system 50 monitors the patient 40 and outputs a signal indicating the patient's condition. The Auxiliary Patient Monitoring System 50 includes a monitoring device in the form of a cardiopulmonary monitor or respiratory monitor.

The respiratory circuit monitoring system 60 monitors the output of the respiratory support device 30, i.e. the percentage of absorbed oxygen to be distributed to the patient. It outputs a signal representing the monitoring result. The respiratory circuit monitoring system 60 has a device in the form of an oxygen analyzer and additionally a pressure transducer.

The output from the assistant patient monitoring system 50 and the respiratory circuit monitoring system 60 is transmitted to the control device 10. The control device 10 determines the output FiO _{2 value based on the signals and input SpO 2} transmitted from the assisted patient monitoring system 50 and the respiratory circuit monitoring system 60.

As shown in FIG. 2, the control device 10 is a stand-alone device. The control device 10 receives the controller 11 that determines the _{output FiO 2} value based on the _{input SpO 2 value, and the input that receives the signal representing the input SpO 2} value and outputs the signal representing the determined output FiO ₂ value. / Has an output interface 12. The input / output interface 12 also receives inputs from the assisted patient monitoring system 50 and / or the respiratory circuit monitoring system 60.

The controller 11 is in the form of an electronic controller that includes one or more digital microcontrollers or microprocessors that perform each step of the method described herein. The controller 11 has one or more application-specific integrated circuits and / or field programmable gate arrays configured to perform each step of the method.

The control device 10 further includes a memory 13 for recording the _{received input SpO 2 value.} The memory 13 stores machine-readable instructions that define the steps of the method described herein and is read and executed by the controller 11 to perform one or more steps of the method.

The control device 10 also has a user interface display 14 that displays a user interface that presents various information to the user (for example, a bedside caregiver) and receives instructions input by the user through the user interface. The received user input is transmitted to the controller 11.

The control device 10 has a data collection device (DAQ) 15, which collects signals / data transmitted from other components of the system 100.

The oximeter 20 has a pulse oximeter 21. The pulse oximeter 21 _{measures SpO 2} of the patient 40 and transmits an output signal representing the _{SpO 2} value to the control device 10. The pulse oximeter 21 further comprises (1) a perfusion index, which is a pulsatile metric of an oxygen measurement waveform with a small value potentially associated with a _{spurious SpO 2 value, and (2) a SpO 2} plethysmograph waveform (Pleth). To measure. The output signal representing the perfusion index and the output signal representing the SpO ₂ plethysmograph waveform are transmitted from the pulse oximeter 21 to the control device 10. The pulse oximeter 21 further measures the heart rate (HR _{Pleth ) derived from the SpO 2} plethysmograph waveform and transmits it to the control device 10.

The respiratory support device 30 has an air-oxygen mixer 31. A signal representing the determined output FiO ₂ value is directed from the controller 10 to a servomotor 32 custom-installed in the air-oxygen mixer 31 thereby a mixer FiO ₂ selection dial via an annular gear mechanism. Can be automatically rotated. The servomotor 32 and gear system have sufficient torque and accuracy and can make _{fine adjustments to FiO 2} (eg, minimum ± 0.5%) accurately and repeatedly. The servomotor 32 may have a low hold torque so that the mixer dial can be turned manually. This manual intervention is detected by the position sensor and results in a switch to manual mode in which _{FiO 2 is no longer automatically controlled.}

At the beginning of the FiO ₂ automatic control, the servomotor calibration is checked and modified as needed. Servo motor calibration can be periodically checked and / or changed during long-term use if desired.

The control device 10 uses the feedback signal from the servo motor 32 at _{the position of the FiO 2} selector dial (servo FiO _{2 ) to confirm that the change of the output FiO 2} value is correctly executed by the servo motor 32.

As shown in FIG. 2, the assisted patient monitoring system 50 includes a respiratory monitor 51 and a cardiopulmonary monitor 52. The respiration monitor 51 monitors the respiration of the patient 40 and outputs a signal indicating the respiration rate of the patient 40. The cardiopulmonary monitor 52 monitors the electrocardiogram (ECG) of the patient 40 and outputs a signal representing _{the heart rate (HR ecg) derived from the electrocardiogram monitoring.} The output from the respiratory monitor 51 and the cardiopulmonary monitor 52 is transmitted to the control device 10.

The respiratory circuit monitoring system 60 includes an oxygen analyzer 61 and a pressure transducer 62. The oxygen analyzer 61 monitors the output of the air oxygen mixer 31, i.e. the mixed gas to be distributed to the patient 40, _{and outputs a signal representing the measurement FiO 2} to the control device 10. The pressure transducer 62 converts the pressure in the suction rim of the CPAP circuit and outputs a signal representing the CPAP circuit pressure to the control device 10.

_{Confirmation that the change in output FiO 2} sent from controller 10 was correctly executed by the automatic air-oxygen mixer 31 was made in the measurement _{of output FiO 2 from the air-oxygen mixer 31 using the oxygen analyzer 61.} Is based. The information from the oxygen analyzer 61 (measurement FiO ₂ ) is digitized from the analog signal and further selected flow time delay (which can be 5 seconds or any other suitable value, selected at the time of system setup). The gas flow time and the downstream equilibrium of the mixer are compensated by being offset by (or determined).

Based on these input signals from the oximeter 20, the assisted patient monitoring system 50 and the inspired oxygen monitoring system 60, the control system 10 determines the _{output FiO 2 value.}

_{Inputs other than the inputs SpO 2} from the oximeter 20, the assisted patient monitoring system 50 and the respiratory circuit monitoring system 60 are referred to as additional inputs. As mentioned above, additional inputs include measured FiO ₂ , CPAP circuit pressure, respiratory rate, perfusion index, plethysmograph waveform, HR _Pleth and HR _ecg .

The control device 10 further includes an alarm unit controlled by the controller 11 to trigger an audio and / or visual alarm. For example, _{if deviations from either the Servo FiO 2} or the Measured FiO ₂ output FiO ₂ values exceed the permissible limits (1 and 2%, respectively), an alarm is triggered. In one example, a high level alarm and a switch to manual mode are performed for each occurrence of a deviation of 5 and 10%, respectively.

Alternatively, the control device 10 may be mounted or integrated with another device rather than a stand-alone device. For example, the control device 10 may be integrated within the oximeter 20 or within the respiratory support device 30.

As described in detail below, the controller 11 has a feedback-type core component adapted and configured for automatic oxygen control for preterm infants. The feedback controller has a mechanical, digital and / or electronic circuit for generating an output control signal based on an input control signal and an internal control value (also called a term) in the feedback controller. The internal control value is
(A) Immediate control value that adjusts the output based on the current value of the input,
(B) Cumulative control value that adjusts the output based on the preceding or past value of the input,
(C) Predictive control value that adjusts the output based on the predicted future value of the input,
Including the total of.

How to automatically control inhaled oxygen distribution
(A) A step of receiving signals representing a plurality of input oxygen saturation (SpO _{2) values for the patient, and}
(B) A step of generating a control value based on the _{input SpO 2} value and the target SpO _{2 value, and}
(C) It comprises a step of generating an output inspired oxygen concentration (FiO ₂ ) value based on a controlled value and a reference inspired oxygen concentration (rFiO _{2) value.}

The control value includes an immediate control value associated with a comparison between the _{current input SpO 2} value and the target SpO _{2 value.} The immediate control value is _{generated based on the input SpO 2} value, the target SpO ₂ value, and the immediate gain coefficient.

The control value further includes a cumulative control value generated based on the cumulative relationship between the _{input SpO 2} value and the target SpO _{2 value.} Cumulative relationship between the input SpO ₂ values and the target SpO ₂ value, the difference may be accumulated between the input SpO ₂ values and the target SpO ₂ value. The cumulative control value is _{generated based on the input SpO 2} value and the target SpO ₂ value, and may be adjusted by the cumulative gain coefficient.

The control value further includes a predictive control value generated based on the predictive relationship between the _{input SpO 2} value and the target SpO _{2 value.} The predictive relationship may be the time derivative of the difference between the _{input SpO 2} value and the target SpO _{2 value.} The predicted control value is _{generated based on the input SpO 2} value, the target SpO ₂ value, and the predicted gain coefficient.

Feedback controller, Immediate control value, the accumulated control value, and generates a control value based on the predicted control values and RFIO ₂ value, and determines an output FiO ₂ values.

In the feedback controller, an error is defined as a deviation of the process signal from the setpoint. The feedback controller may be a proportional integral differential (PID) controller.

The PID feedback controller is enhanced in many ways. Measurements of the severity of pulmonary dysfunction are periodically obtained by automatic assessment of current oxygen requirements. Immediate control enhancements include severity of pulmonary dysfunction, error attenuation within the target range, and modulation based on error upper bounds during hypoxia. The enhancement of the cumulative control value includes the upper limit of the magnitude _{of the integrand, the compensation of the non-linear PaO 2-} SpO ₂ relationship, and the limitation of the increase of the integrator in the room air.

For the PID controller, the value of the manipulated signal output at each moment is proportional to the error, its integral and its derivative, and in each case a different multiplication factor, namely the immediate gain factor, the cumulative gain factor, and. It has a predicted gain coefficient (referred to as K _p , K _i , and K _d , respectively). For the PID Fordback controller, the immediate control value is also called the proportional term, the cumulative control value is also called the integral term, and the predictive control value is also called the differential term. These three are called PID terms.

In the method for automatically controlling inhaled oxygen distribution described herein, the immediate control value is the error value associated with the difference between the _{current input SpO 2} value and the target SpO _{2 multiplied by the immediate gain factor.} It is generated by doing. The error value is generated by determining the error (e), i.e., _{the difference between the current input SpO 2} value and the target SpO _{2 value.} Alternatively, the error value may be generated by any other suitable mathematical method that compares the _{current input SpO 2} value with the target SpO _{2 value.}

For the PID controller, _{the input value of SpO 2} (assuming a valid signal) and the midpoint of the selected target range (for example, if the target range is 91 to 95%, the midpoint is 93%). The numerical difference between them may be used as an error (e).

Further, the cumulative control value is generated by multiplying the cumulative difference between the _{input SpO 2} value and the target SpO _{2 value by the cumulative gain coefficient.} For example, the cumulative control value is generated by multiplying the sum of the error values by the cumulative gain factor for the digital signal, or by multiplying the integration of the error values by the cumulative gain factor for the analog signal. .. Alternatively, the cumulative control value may be generated by other suitable mathematical methods that give rise to a cumulative relationship between the _{input SpO 2} value and the target SpO _{2 value.}

For PID controllers, the integrand (∫edτ) is the sum or integral of all errors (subject to the limitations described below). The integral term in PID control serves the advantage of overcoming certain state errors.

Further, the predictive control value is a time-division difference between continuous error values (for a digital signal) or a derivative of the _{error value (that is, the difference between the input SpO 2} value and the target SpO _{2 value).} It is generated by multiplying the coefficient (for an analog signal) by the predicted gain coefficient. Alternatively, the predictive control value may be generated by other suitable mathematical methods that give rise to a predictive relationship between the _{input SpO 2} value and the target SpO _{2 value.}

For the PID controller, the derivative (de / dt) is the SpO ₂ gradient by linear regression over the preceding 5 seconds, which provides prediction of future errors in PID control.

For example, the sum of each of the PID terms _{is expressed as ΔFiO 2} (indicated by equation (1) below).

As mentioned above, the output FiO ₂ value (FiO 2 to be distributed to the patient) is determined based on the control value and the reference inspired oxygen concentration (rFiO _{2) value.}

In certain embodiments, the output SpO ₂ value is the sum of the _{corresponding control value and the corresponding rFiO 2} value, as shown in equation (2) below.

For example, FiO ₂ is _{the sum of ΔFiO 2} and rFiO ₂ (indicated by equation (2) below). In addition, FiO ₂ is rounded to ± 0.5% and forced to a value between 21 and 100%. That is, any value less than 21% is rounded up to 21%, and any value greater than 100% is rounded down to 100%.

The rFiO ₂ value represents the current baseline oxygen requirement, which indicates the severity of the patient's pulmonary dysfunction. It may be a predicted value or a range of values or may be selected by user input. For example, the rFiO ₂ value is predetermined as a number between 21% and 60%, or any other suitable number up to 100%. The rFiO ₂ value has an initial value and is periodically modified over time, as described in detail below. The time interval for repeating the determination is fixed (eg, any interval of 30 minutes to 2 hours) or is not alternatively determined. Thus, it becomes possible to detect and respond to gradual changes in base oxygen requirements that occur in patients with respiratory dysfunction.

In certain embodiments, the immediate gain factor has an initial value chosen between -2 and -0.2, such as -1.

In certain embodiments, the cumulative gain factor has an initial value chosen between -0.25 and -0.005, for example -0.0125.

In certain embodiments, the predicted gain factor has an initial value chosen between -2 and -0.25, for example -1.

Each value of K _p , K _i , and K _d is determined based on a predetermined reference value or value range. For example, K _p , _Ki , and K _d are determined based on reference values or a range of values derived from simulation studies using data from preterm infants. Each value of the coefficient may be negative, which means that the PID terms work together to correct the error. Exemplary ranges of coefficient values, _{K p} is -0.2 from -2 _{K i} is -0.005 from -0.25 a -0.25 a _{K d} -2, for example, is _{K p} -1, _{K i} is -0.0125, _{K d} is -1. As explained in detail below, the fixed value of K _p may be modified in accordance with the seriousness of pulmonary dysfunction, it is further improved through the self-regulating process during the automatic control, for example, 30 to every 60 minutes (Or, for example, another suitable time interval sufficient for a suitable evaluation of 10 minutes or more and 120 minutes or less).

In certain embodiments, the immediate control value may be modified.

In certain embodiments, the method is
It further has a step of determining the _{target SpO 2} value based on the target SpO _{2 range.}
When the current input SpO ₂ value is _{within the target SpO 2} range, the attenuator is applied to the immediate gain factor.
The attenuator is generated based on the current input SpO ₂ value and the midpoint of the _{target SpO 2 range.}

The attenuator is a multiplier of the ratio proportional to the difference between the _{current input SpO 2} value and _{the midpoint of the target SpO 2 range.}

Further, when the input SpO ₂ values of the current is smaller than the SpO ₂ value of the target, an error value related to the difference between the input SpO ₂ values and the target SpO ₂ value of the current is, the upper limit specified in the maximum difference is selected NS.

For PID control, the determination of the proportional term is modified if _{the input SpO 2} value is _{within the target SpO 2 range.}

System 100 targets the midpoint of the target range and defines any deviation from this value as error (e). _{In recognizing that any SpO 2} value within the target range is acceptable, the error associated with deviations from the midpoint of the target range is reduced by the _{multiplier K pfm proportional to the distance from the midpoint (target).} Range attenuation). For example, for a target range having a span of ± 2 from the midpoint (eg, 91-95%), a multiplier K _{pfm of} 0.25 to | e | = 1 is applied to K _p , and | e | = A multiplier of 0.5 to 2 K _pfm is applied.

Also, if _{the relative inaccuracy of SpO 2} monitoring is given with a value of 80% or less, negative errors are capped, for example, to 15% for the determination of the proportional term.

In certain embodiments, the cumulative control value may be modified.

In certain embodiments, non-linear weighting compensation may be applied to the cumulative control value based on a non-linear, predetermined relationship between the partial pressure of arterial oxygen (PaO ₂ ) and SpO _2.

Further, the cumulative control value may be modified so that the control value is capped at the selected maximum control value.

The integration term may be modified for PID control.

In the case of constant hypoxia, the _{maximum ΔFiO 2} output from the PID controller is a user-set value (that is ± 30-40%, i.e., in _{recognition that the integral term increments FiO 2 more and more.} The size of the integrand is limited to 30-40% above or below rFiO _{2). In hyperoxia (ie, SpO 2} beyond the target range during assisted oxygen) following a hypoxic event as an overshoot, an error with a high SpO ₂ value is similar _{to a PaO 2 deviation from the tolerance.} Not proportional to (ie, non-linear PaO _2- SpO ₂ relationship). Thus, as long as the integrand remains negative (ie, _{the tendency to increase ΔFiO 2} ), the error multiplier applies to positive errors proportional to the _{associated ΔPaO 2 value.} In one embodiment, the corrected error is added to the integrand by each iteration, while the integrand remains negative. The error multipliers are shown in Table 1.

The error multiplier has the effect of returning the rapidly increasing negative integrand to zero, thus mitigating the overshoot.

Further, in a certain embodiment, generating a cumulative control value means that (i) the current output FiO ₂ value is at the indoor air level and (ii) the current input SpO ₂ value exceeds the target SpO _{2 value.} Sometimes it involves limiting the increase in cumulative control values.

If the integrand is positive (ie, ΔFiO ₂ tends to decrease), additional positive errors are added to the integrator only while the set of _{FiO 2 remains above room air (21%).} NS. In the case of room air (ie FiO ₂ = 21%), _{continuous SpO 2} values above the target range are no longer considered to represent continuous hyperoxia and positive errors are no longer added to the integrand. .. That is, these positive error values can be null or zero. This avoids the establishment of a positive integrand that delays the appropriate response from the integral term for the next onset of hypoxia.

In certain embodiments, the predictive control values may be modified.

In certain embodiments, the predictive control value is null if the input SpO ₂ value is above the selected SpO _{2 threshold during all negative SpO 2} gradient determination periods.

The differential term may be modified for PID control. For example, the derivative term may be modified during hyperoxygenosis.

In some embodiments, if all of the last five SpO ₂ value exceeds the setpoint (hyperoxia), negative SpO ₂ gradient becomes null. Therefore, the pressure increase _{of ΔFiO 2} due to the differential term is avoided in hyperoxygenosis.

FIG. 3 shows process 300 performed by controller 11 to generate proportional, integral, and derivative terms, including modifications as described below.

As shown in FIG. 3, in step S302, the value of error (e) is determined as a numerical difference between the _{input SpO 2} value and the target SpO _{2 value, as shown in Equation 3 below.}

Next, in step S304, the proportional term is modified using the following steps.

(A) Based on the error value, the value of the multiplier K _pfm is selected as follows.
If | e | ≤ 1 (ie, the error is less than or equal to 25% of the target range, and therefore the input SpO ₂ value is within the target range and is _{close to the target SpO 2} value), then K _pfm = 0.25,
Alternatively, if | e | ≤2 (ie, the error is greater than 25% of the target range but less than or equal to 50% of the target range, therefore the input SpO ₂ value is not close to the target SpO _{2 value.} , Within the target range), K _pfm = 0.5,
In other cases (ie, the error is greater than 50% of the target range and therefore the input SpO ₂ value is outside the target range), K _pfm = 1.

_{(B) CPAP circuit Adjust K pfm} as follows based on pressure and respiratory rate.
If CPAP circuit pressure = low, then K _pfm = 2 * K _pfm (ie, a reduction in circuit pressure leads to _{double K pfm),}
Alternatively, if the respiratory arrest period is 5 to 15 seconds, then K _pfm = 2 * K _pfm for 30 seconds (ie, respiratory arrest results in _{double K pfm for 30 seconds).}

(C) During hypoxia, the proportional term error upper limit is applied as follows.
If e> -1.5% (ie, the patient is hypoxic), the proportional term error e _p = -15% (ie, caps the proportional term error),
Others, e _p = e

(D) The proportional term is calculated as follows.
Proportional term = P (t) = K _pfm * K _p * e _p

After modifying the proportional term, the logic moves to step S306 and modifies the integral term using the following steps.

(A) To a non-linear, pre-determined relationship that can be pre-determined using well-known relationships between blood oxygen levels and _{high SpO 2 values, including those described by the Severing House equation.} based on, it is determined as follows nonlinear compensation multiplier (K _s).
If the preceding integral term I (t-1) <0 (ie, the ratio integral function remains negative) and e> 0 (ie, SpO ₂ exceeds the target range).
If SpO ₂ = 92, then K _s = 1.2,
Besides, if SpO ₂ = 93, K _s = 1.4,
Besides, if SpO ₂ = 94, K _s = 1.7,
Besides, if SpO ₂ = 95, K _s = 2.2,
Besides, if SpO ₂ = 96, K _s = 2.9,
Besides, if SpO ₂ = 97, K _s = 4.4,
Besides, if SpO ₂ = 98, K _s = 7.9,
Besides, if SpO ₂ = 99, K _s = 20.1,
Besides, if SpO ₂ = 100, K _s = 50,
Others _{apply K s} = 1 (ie, apply the nonlinear compensating multiplier to positive errors proportional _{to the associated ΔPaO 2 value)}

(B) The increase of the integrand in the room air is limited as follows.
If FiO ₂ = 21% (ie, room air) and e> 0, then dI = 0 (ie, positive errors are not added to the integrand).
_{Others, dI = K i * K s} * e

(C) The integral term is determined as follows.
Integral term = I (t) = I (t-1) + dI

(D) The upper limit of the size of the integrand is specified as follows.
If | I (t) |> | ΔFiO 2 of the maximum value / _{K i} | if, I (t) = (sign) * (maximum value _{/ K} i of DerutaFiO ₂₎ (i.e., ΔFiO ₂ values selected Upper limit of the value of the integral term based on)

After modifying the integral term, the derivative term is modified in step S308 using the following steps.

(A) The differential term is evaluated as follows.
Derivative term = D (t) = K _d * de / dt, where de / dt is determined by linear regression over 5 seconds.
(B) Null for negative gradients and hyperoxygenosis as follows.
If de / dt <0, (SpO ₂ (t)> target SpO ₂ ), (SpO ₂ (t-1)> target SpO ₂ ) (SpO ₂ (t-2)> target SpO ₂ ), (SpO ₂₎ If (t-3)> target SpO ₂ ) and (SpO ₂ (t-4)> target SpO ₂ ), (that is, _{all of the five recent SpO 2} values exceed the set point. )
D (t) = 0 (ie, the negative SpO ₂ gradient is null)

Further, the control value may be further generated based on the _{rFiO 2 value.}

In certain embodiments, the immediate control value (proportional term) is further based on the _{rFiO 2 value.}

The immediate control value (proportional term) is modified by the modification value determined from _{the rFiO 2 value.} The modification value is _{determined using the monotonic relationship with rFiO 2} , that is, based on the monotonic function. For example, K _p is modified from a predetermined initial reference value or from its current value by a value determined _{from the rFiO 2 value.} This correction value is, for example, by a scaling factor proportional to the severity of pulmonary dysfunction, as indicated by RFIO ₂ of the current, increasing the effective value of the immediate control values in order to increase the RFIO _2, for example, K _p The fixed value of is multiplied by a factor in the range of 0.5 to 1.5 for _{rFiO 2} of 21% to 60% in the corresponding range. (For example, for 21% rFiO ₂ , the scaling factor is 0.5, for 40% rFiO ₂ , the scaling factor is 1.0, and rFiO ₂ changes from 21% to 60%. The scaling factor changes linearly from 0.5 to 1.5 in proportion to). Alternatively, the scaling factor may be performed as an equivalent modification that modifies the immediate control value. The addition of K _p in this way, to compensate for the inverse relationship between the severity of gain and pulmonary dysfunction.

Also, the method is
During the performance analysis time interval, the process of receiving signals representing _{multiple input SpO 2 values, and}
The process of generating performance evaluation results based on the _{input SpO 2} values received during the performance analysis time interval, and
It further has a step of generating a control value based on the performance evaluation result.

In addition, the immediate gain coefficient is modified based on the performance evaluation result.

In some embodiments, the values of K _p is repeatedly corrected during the automatic control of the inspired oxygen distribution. The performance analysis of the automatic control of inhaled oxygen distribution _{is performed cyclically based on the input SpO 2} values received over the performance analysis time interval to generate performance evaluation results, where the K _p value is the performance evaluation result. It will be modified based on.

In certain embodiments, the performance evaluation results show that the input SpO ₂ value within the performance analysis time interval is in the hypoxia range, and the input SpO ₂ value within the performance analysis time interval is in the hypoxia range. It is produced based on at least one of a certain excess oxygen time intervals.

Also, in certain embodiments, performance assessment results are generated based on the ratio of the excess oxygen time interval to the low oxygen time interval.

Moreover, in certain embodiments, the method
It further has a step of determining the _{target SpO 2} value based on the target SpO _{2 range.}
The performance evaluation result is
_{The input SpO 2} value within the performance analysis time interval is the target time interval within the target SpO _{2 range, and the input SpO 2} value within the performance analysis time interval is the eupoxia time interval within the eupoxia range. , the input SpO ₂ values in a range exceeding the target SpO ₂ range in the target SpO ₂ within or room air, is generated based on at least one Eupokishia time interval where.

The performance time interval is a fixed time interval, eg, predetermined or set by the user. The performance analysis time interval may be 60 minutes and the analysis is performed based on _{SpO 2 data recorded within the last 60 minutes prior to the analysis.} Alternatively, the performance analysis time interval may be a variable time interval based on the analysis results or instructions entered by the user (eg, any time interval between 30 minutes and 2 hours).

The analysis is performed at regular intervals. For example, the analysis is performed once every 30 minutes. The analysis may be performed continuously or once after a time interval, eg, any suitable time interval of up to 2 hours. The frequency of analysis may be set by the user.

The analysis begins with hypoxia (SpO ₂ <85%) and continues at a time interval (any selected suitable time interval between 2 and 10 minutes) that exceeds its resolution, all within the time window. Based on the response to the event.

The total time of hypoxia (80 to 84% of SpO ₂₎ and severe hypoxia (SpO ₂ less than 80%), as well as, SpO of hyperoxia when receiving oxygen into (97-98%) ₂ Time intervals for overshoot and severe hypoxia (99-100%) are measured as described with reference to FIG. 4 below.

From these data, a weighted performance factor is derived as the ratio of the time of hyperoxygenosis to the time of hypoxia, the values being less than 1 and greater than or equal to 1, respectively, underpower _Kp and The K _p of overpower is shown. Current values of K _p is varied until 10% ± every 30 minutes as a result of this analysis.

Also, _{the ratio of time that SpO 2} was within the target range and within the eupoxia range (SpO ₂ exceeds the target range in the target range or indoor air), and the occurrence of hypoxia and hyperoxygenosis in oxygen. The rate is calculated.

In certain embodiments, the performance assessment results do not match certain conditions that are adjusted by the user and set in the range of 50-80%, including that the controller output is substantially below the minimum requirement for the eupoxia ratio. Triggers an alarm.

FIG. 4 shows a process 400 executed by the controller 11 that modifies _Kp based on the performance evaluation results described below.

As shown in FIG. 4, process 400 determines in step S402 whether 30 minutes have passed since the occurrence of automatic control or the prior performance analysis.

If the determination result is no, the process 400 ends. If it is determined that 30 minutes have passed, process 400 moves to step S404 and uses the following steps to perform a performance analysis based on control performance over the preceding 60 minute time window. Run.

(A) Calculate the percentage of time in the range of eupoxia, hypoxia and hyperoxygenosis.
t _{severe hypoxia} : SpO ₂ <80%
_hypoxia : 80% ≤ SpO ₂ ≤ 84%
_teupoxia : _{SpO 2} , FiO ₂ = 21% within or above the target range
_hyperoxia : 97% ≤ SpO ₂ ≤ 99% when oxygen is received
t _{severe hyperoxia} : 99% ≤ SpO _{2 when oxygen is received}
(That is, quantify the total time of hypoxia, severe hypoxia, hyperoxia, severe hyperoxygenosis and eupoxia)

(B) The weighted performance factor is calculated as follows.
_{C performance = (t severe hyperoxia +} t hyperoxia) / (t hypoxia + t severe hypoxia)
(I.e., an oxygen time ratio hyperkinesia, excess K _p, respectively shortages and forces including severe hyperoxia against time hypoxia including severe hypoxia.)

(C) The new K _p is calculated as follows.
If C _performance ≤ 0.7, then K _p = K _p * 1.1
If 0.7 <C _performance ≤ 0.85, then K _p = K _p * 1.05
If 1.15 ≤ C _performance ≤ 1.3, then K _p = K _p * 0.95
If 1.3 ≤ C _performance , then K _p = K _p * 0.9
(I.e., change the value of K _p based on the ratio of the time)

(D) Eupoxya time is calculated as follows.
Eupoxya time = ( _teupoxya × 100) / t _total
(That is, _{the percentage of time SpO 2} was in the eupoxia range)
If eupoxia time <target range approaching the target, an alarm is issued that target range approach = truth. (That is, an alarm is triggered when automatic control falls below the minimum requirement)

Also, the method for automatically controlling the inhaled oxygen distribution is
(A) A step of generating an _{rFiO 2} evaluation result based on the _{input SpO 2 values over the rFiO 2} evaluation time interval and the respective output FiO _{2 values.}
It further comprises a step of modifying the _{rFiO 2} value based on the _{rFiO 2 evaluation result.}

The rFiO ₂ value has an initial value and is repeatedly modified over time.

In certain embodiments, the analysis is performed on a regular basis. For example, the analysis is performed once every 30 minutes, which is called the evaluation time frequency. The analysis is performed once after a time interval of any suitable time between 30 minutes and 2 hours. In the evaluation time frequency, between the _{set FiO 2} and SpO ₂ within the shifting time window (evaluation time interval), assuming that the fixed V / Q ratio and variable shunt caused disturbance of oxygen supply. Relationship analysis is performed. The time window is 60 minutes or any suitable time interval (arbitrarily selected from 30 minutes to 2 hours). The V / Q value is derived, for example, using well-known methods and formulas for calculating the V / Q ratio, and then the rFiO ₂ value is modified to overcome its effect on oxygen supply. This will be _{the new value for rFiO 2} , which will be forced within ± 10% of the preceding value. Therefore, abrupt changes in _{rFiO 2 can be avoided.}

FIG. 5 shows a process 500 _{performing the rFiO 2} value modification performed by the controller 11.

As shown in FIG. 5, process 500 determines in step S502 whether automatic control has been started or 30 minutes have passed since _{the latest time reference FiO 2} (rFiO _{2) was updated.}

If the result of the determination is no, the process 500 ends. If it is determined that 30 minutes have passed since the automatic control was started or the last time reference FiO ₂ (rFiO ₂ ) was updated, process 500 moves to step S504 and sets the rFiO ₂ value below. Update using the steps in.

(A) Sliding window analysis The sliding window analysis of FiO ₂ and SpO ₂ is executed for 60 minutes to obtain a V / Q ratio, and rFiO ₂ is obtained from the V / Q ratio.

(B) rFiO ₂ is forced within ± 10% of the preceding value as follows.
If (rFiO ₂ -preceding rFiO ₂ ) / (preceding rFiO ₂ )> 0.1, then
rFiO ₂ = leading rFiO ₂ + (sign) (rFiO ₂ -preceding rFiO ₂ ) * 0.1 * leading rFiO ₂
(That is, it determines _{a new rFiO 2} value forced within ± 10% of the preceding value, thus avoiding abrupt changes in _{rFiO 2).}

(C) At the start of automatic control, _{the current value of FiO 2} or the value entered by the user is used as the initial value of the _{rFiO 2 value.}

Also, the method for automatically controlling the inhaled oxygen distribution is
By classifying into one of a plurality of verify levels in the hierarchical verification process input SpO ₂ values of (a) current, and generating an SpO ₂ verification result based on the input SpO ₂ values of the current,
(B) Further includes a step of determining the _{output FiO 2} value based on the _{SpO 2 verification result.}

The following hierarchical verification levels are adopted.
(A) Level I: If the SpO ₂ input matches the first condition, it corresponds to the absence of the _{SpO 2 input.}
(B) Level II: If the SpO ₂ input matches the second condition, it corresponds to the suspicion of _{the SpO 2 input.}
(C) Level III: If the SpO ₂ input matches the third condition, it corresponds to the invalid _{SpO 2 input.}

Also, in certain embodiments, the method
SpO ₂ A signal representing the heart rate derived from the plethysmograph waveform,
It further comprises the step of receiving at least one of a heart rate signal and a perfusion index derived from ECG monitoring.
Verification of the current input SpO _{2 value is}
Received heart rate derived from SpO ₂ plethysmograph waveform, and received heart rate derived from electrocardiogram monitoring,
Judgment is based on at least one of the received perfusion indices.

For _{verification of the SpO 2} signal, some or all of the following incidental inputs are supplied as digital signals.
(A) Heart rate derived from _{SpO 2 plethysmograph waveform (HR press} )
(B) Heart rate derived from ECG monitoring (HR _ecg )
(C) Perfusion index, an oximetry waveform pulsatile metric with a low value potentially associated with a _{false SpO 2 value.}

At the start of automatic control, and for about 24 hours thereafter (perfusion check time interval), a perfusion index value representing optimal perfusion is recorded when the plethysmograph waveform is stable and the signal is strong. The perfusion check time interval may be any suitable time interval, including any arbitrarily selected time interval of 6 hours to 2 days.

FIG. 6 shows a process 600 performed by the controller 11 to determine the perfusion index value.

As shown in FIG. 6, process 600 determines in step S602 whether automatic control has been started or 24 hours have passed since the latest perfusion index was observed. If the determination result is no, the process 600 ends. If it is determined that automatic control has been started or 24 hours have passed, treatment 600 proceeds to step S604 and a new optimal perfusion index value, eg, the 95th in the last 24 hours. Enter a point.

Pretismograph waveforms, plethysmograph waveforms recorded over the last 10 seconds (or any suitable time interval, including any selected time interval of 5 to 20 seconds), are digitized by an analog-to-digital converter. Is input as an analog signal. Digital and analog signals may be acquired using a data acquisition device.

_{The waveform analysis verifies SpO 2} by analyzing the plethysmograph signal from the pulse oximeter and confirms that it matches the characteristics expected from the effective plethysmograph signal. Evaluation methods include analysis of statistical characteristics of signals (eg, mean and variance), classical signal processing techniques (eg, autocorrelation), logic algorithms (including fuzzy logic), and pattern recognition techniques (including neural networks). ), Or a combination.

For example, _{an exemplary process for performing SpO 2} plethysmograph analysis involves the following steps:
(A) A step of periodically acquiring effective representative value plethysmograph tracking from individual patients.
(B) The _{process of standardizing the current input SpO 2} plethysmograph signal in the two axial directions, and as a result, the process in which the period and peak correspond to the representative plethysmograph tracking throughout the amplitude, and (c) a plurality of straight lines. The process of comparing the two signals by regression with a mean square error that gives an indicator of the deviation of the current signal from the _{representative SpO 2 plethysmograph waveform.}

Alternatively, the two signals may be compared by other methods including pattern recognition, such as linear discriminant analysis or artificial neural networks. Using these methods, a pre-record database of plethysmograph waveform recordings with _{SpO 2 classified as valid or invalid, such that the monitored SpO 2} signal is used to classify as valid or invalid, Used for pattern recognition training and validation.

Along with these additional inputs, the following hierarchical verification method is adopted.
(A) Level I: When the SpO ₂ value is not zero or numerical value (b) Level II: When the waveform does not match the general plethysmograph waveform after normalization in both axial directions (c) Level III:
i. Is the perfusion index less than 30% of the optimal value (or any other suitable value, eg, any value between 10% and 50%) and the waveform is suspicious?
ii. The absolute value of HR _{Pleth- HR ecg} is greater than 30 bmp (or any other suitable value, eg, any value between 20 bpm and 50 bpm) and the waveform is suspicious, or iii. It plunged SpO ₂ (e.g., greater than 15% at 5 seconds, or any other suitable values representative of the sudden and deep fall of SpO _2, wherein, since often sudden fall of SpO ₂ is questionable readings, falling Is present for a period of time, eg, 5 seconds), and is accompanied by either a suspicious waveform, a heart rate discrepancy, or a perfusion index discrepancy (indicating suspicious hypoxia). )case.

FIG. 7 shows an example of the hierarchy verification process 700 executed by the controller 11.

As shown in FIG. 7, in step S702, _{it is determined whether or not the input SpO 2} value is zero or non-numerical. If yes, level 1 of the flag is set as true (step S704) and process 700 ends.

If the input SpO ₂ value is not zero or non-numeric, then level 1 of the flag is set as false (step S706) and in step S708 the input SpO ₂ signal is tested for matching with the plethysmograph waveform. If, after both axial normalizations, _{it is determined that the waveform of the input SpO 2} signal does not match the general plethysmograph waveform, then level 2 of the flag is set as true (step S712), ie SpO _2. The input will be suspicious. If the waveform of the input SpO ₂ signal matches the general plethysmograph waveform, level 2 of the flag is set as false.

Next, the logic proceeds to step S714, determines whether or not the perfusion index is less than 30% of the optimum value, and if the perfusion index is less than 30% of the optimum value, the PI mismatch of the flags is true. Is set as (step S716), or the PI mismatch of the flags is set as false (step S718).

Next, in step S724, a test for determining whether or not the absolute value of _{HR Pleth- HR ecg exceeds 30 bmp (step S720) is executed.} If yes, the flag HR mismatch is set as true (step S722), and if no, the flag HR mismatch is set as false (step S724).

Further, in step S726, it is determined whether or not the level III condition is satisfied. For example, the following logic is used to _{determine whether the SpO 2} value is invalid.
(PI mismatch and level 2) or (HR mismatch and level 2) or ((SpO ₂ decreases by more than 15% in 5 seconds) and (level 2 or PI mismatch or HR mismatch))

If the result is positive, the level III flag is set as true (step S728). That is, the SpO ₂ input is invalid. If the result is not positive, the level III flag is set as false (step S730).

In certain embodiments, when the input SpO ₂ value is determined to be invalid (eg, a missing or invalid signal event), FiO ₂ takes the output SpO ₂ value at the current value, i.e., the previously recorded FiO ₂ value. Set to retain. In a persistent missing or invalid signal event, the output FiO ₂ value is directed towards _{rFiO 2 beyond the alarm trigger described below.}

FIG. 8 shows a process 800 performed by the controller 11 _{for determining the output FiO 2} value based on the validity of _{SpO 2.}

As shown in FIG. 8, a test is performed in step S802 _{to determine if the input SpO 2} value is missing or invalid. If the input SpO ₂ values when it is determined that the missing or invalid, the output FiO ₂ value is set equal to the output FiO ₂ values preceding in step S806. If in step S802 it _{is determined that the input SpO 2} value is not missing or invalid, the logic moves to step S804, where the output FiO _{2 value is the input SpO 2} value, eg, using the following steps. Determined based on.
(A) ΔFiO ₂ calculated _{ΔFiO 2 = P (t) +} I (t) + D (t)
(B) Calculate the _{output FiO 2 value FiO 2} = ΔFiO ₂ + rFiO ₂
(C) If apnea exceeds 15 seconds, change the _{output FiO 2} value by 5% for 30 seconds beyond the apnea interruption _{(d) Round the output FiO 2} value to ± 0.5% (d) e) If FiO ₂ <21%, then output FiO ₂ value = 21%, _{if FiO 2} > 100%, then output FiO ₂ value = 100%.

In certain embodiments, SpO ₂ is disabled during a certain time interval (eg, 30 seconds to prevent continuous missing of signals from babies with _{low SpO 2, or any other suitable time interval).} When determined to be, an alarm (eg, audio and / or visual alarm) is activated. For example, when an alarm is activated with an oximeter probe and an error message to check the connection, the alarm is a voice alarm and the volume of the alarm increases every few minutes (eg, 2 minutes).

In certain embodiments, _{as soon as effective SpO 2} is detected, the alarm is stopped and determination of the immediate control value, cumulative control value, predictive control value, and output FiO ₂ value based on reference inspired oxygen concentration is resumed. Will be done.

Also, the method of automatically controlling the inhaled oxygen distribution is
(A) Step of receiving a signal indicating the respiratory rate,
(B) It further comprises a step of generating an immediate control value based on the respiratory rate.

In certain embodiments, respiratory rate is recorded by an abdominal capsule linked to a respiratory monitor, from which digital pulses per spontaneous breath detected are directed to an automated controller via an analog-to-digital converter.

Onset of respiratory arrest (eg, respiratory interruption of 5 to 15 seconds, or respiratory interruption of any suitable time interval sufficient to predict similar hypoxic events) and apnea (eg, from 15 seconds). Longer or longer than any suitable time interval sufficient to predict similar hypoxia) is identified.

The value of K _p is modified based on this additional input.

For example, respiration pulses, the time interval is greater than the interruption of breathing (e.g., 30 seconds, or any appropriate time interval) produces a 2-fold K _p between. Through this adjustment, the automated controller responds more sensitively to hypoxic events as they occur. If continued until clear apnea respiratory arrest, FiO ₂ is in proportion to the basic K _p value is increased from 2 to 8% temporarily.

In addition, the method of automatically controlling inhaled oxygen distribution is
(A) Step of receiving a signal representing circuit pressure,
(B) It further comprises a step of determining an immediate control value based on the circuit pressure.

For infants receiving continuous positive airway pressure (CPAP) respiratory support, the pressure in the inspiratory rim of the CPAP circuit is converted and input as a digital signal.

For example, a decrease in circuit pressure below a certain percentage of the plateau value (eg, including any selected value between 20% and 50%) _{leads to a double Kp} value of 30. After a second, it triggers an alarm that indicates a decrease in circuit pressure. A complete loss of circuit pressure (<1 cmH ₂ O, ie there is essentially no pressure distribution) triggers a higher level alarm that represents a loss of circuit pressure.

In certain embodiments, the method is
The process of receiving manual override input and
It further comprises a step of determining the _{output FiO 2} value based on a manual override input instead of the control value.

Further, the control device 10 may have an automatic control mode and a manual control mode. The automatic control mode is switched to manual control under some user input (ie, manual override input) and later returned to automatic mode if desired.

For example, the user (eg, bedside staff) can switch the controller 10 to manual mode so that the controller 11 _{does not change the output FiO 2} value and the oxygen supply is completely under manual control. Be placed.

The manual control mode is selected through manual override input in the user interface displayed on the user interface display 14. It is a temporary stop of the controller functions to produce a change in FiO 2 in ₂ (e.g., 30 seconds), or may be either continuous manual operation until the non-selected state.

The manual control mode is a manual mode in which the FiO ₂ selection dial is rotated on an automatic air-oxygen mixer, i.e., triggering a stop for automatic control (eg, a stop for 30 seconds or any selected appropriate time interval). It may be selected by giving an override input.

In the first iteration after the stop, i.e., on the return from manual control mode, the integrand has an output FiO ₂ value that is _{the current value of the servo FiO 2} (ie, user-selected) due to automatic control that is then resumed. ) Is adjusted to be set equal to.

FIG. 9 shows a process 900 executed by the controller 11 for switching between the manual control mode and the automatic control mode.

As shown in FIG. 9, first, in step S902, _{a test is performed to determine whether the servo FiO 2} input has been manually modified, i.e., whether a manual override input has been received. If yes, the manual over-lard mode is initiated in step S904. If the Servo FiO ₂ input was not manually changed, the logic proceeds to step S906 to determine if the manual mode time limit has expired. If the manual mode time limit expires, the control device 10 is set to return to the automatic control mode in step S908. If the manual mode time limit has not expired, process 900 terminates.

Also, one or more alarms are triggered under certain conditions during automatic control.

For example, alarms (eg, audio and / or visual alarms) may be included within the controller 10 _{to reach the rapidly rising maximums of FiO 2 , ΔFiO 2} , and miss _{a SpO 2} signal for more than 30 seconds. Or invalid, warn bedside staff of long-term apnea or loss of circuit pressure for more than 2 minutes, and system malfunction. These alarms may be added, integrated, or replaced with an alarm set in a standard bedside monitor in the NICU.

FIG. 10 shows a process 1000 performed by the controller 11 that controls an alarm based on monitoring the various above-mentioned signals.

As shown in FIG. 10, in step S1002, the process of activating the alarm is executed using the following steps.
(A) If the missing signal = true, the alarm activates the missing signal alarm.
(B) If hypoxia = true, the alarm activates the hypoxia alarm.
(C) If circuit pressure drop = true, the alarm activates the circuit pressure drop alarm.
(D) If circuit pressure loss = true, the alarm activates the circuit pressure loss alarm.
(E) If Servo FiO ₂ Mismatch = True, the alarm activates the _{Servo FiO 2 Mismatch Alarm.}
(F) If Servo FiO ₂ Error = True, the alarm activates the _{Servo FiO 2 Error Alarm.}
(G) If the measurement FiO ₂ mismatch = true, the alarm activates the _{measurement FiO 2 mismatch alarm.}
(H) If the measurement FiO ₂ error = true, the alarm activates the _{measurement FiO 2 error alarm.}
(I) If the target range approach = true, the alarm activates the target range approach alarm.

The process 1000 then proceeds to step S1004 and the alarm is reset using the following steps.
(A) If the alarm (missing signal) = false, the missing signal alarm is canceled.
(B) If the alarm (hypoxia) = false, the hypoxia alarm is canceled.
(C) If the alarm (circuit pressure drop) = false, the circuit pressure drop alarm is canceled.
(D) If the alarm (circuit pressure loss) = false, the circuit pressure loss alarm is canceled.
(E) If the alarm (servo FiO ₂ mismatch) = false, the servo FiO ₂ alarm is canceled.
(F) If the alarm (servo FiO ₂ error) = false, the servo FiO ₂ high alarm is canceled.
(G) If the alarm (measurement FiO ₂ mismatch) = false, the measurement FiO ₂ alarm is canceled.
(H) If the alarm (measurement FiO ₂ error) = false, the measurement FiO ₂ high alarm is canceled.
(I) If the alarm (target range approach) = false, the target range approach alarm is canceled.

After the step S1004 is completed, the process 1000 ends.

Also, as described above, the control device 10 may have a user interface display, which presents various information to the user (eg, bedside caregiver) and is based on the user interface. Displays the user interface that receives the instructions entered by. The received user input is then transmitted to the controller 11.

An example of the user interface 200 displayed on the user interface display 14 is shown in FIG. The user interface 200 includes:
(A) Numerical values SpO ₂ / FiO ₂ display area A01, displaying _{recent input SpO 2} values and recent output FiO _{2 values,}
(B) _{Graphical SpO 2} / FiO ₂ display area A02, which graphically displays trends in _{input SpO 2} values and output FiO _{2 values,}
(C) Oximeter type selection area A03, for the user selecting the type of oximeter used to generate the _{SpO 2 signal.}
(D) _{SpO 2} target range setting area A04, _{for displaying the target SpO 2 in} real time and allowing the user to change it.
(E) Maximum ΔFiO ₂ setting area A05, which indicates the value of ΔFiO 2 in real time and allows the user to change the limit.
(F) The manual control mode button B06, which allows the control device 10 to switch between the automatic control mode and the manual control mode by pressing it,
(G) On / off button B07, which turns on or off the control device 10 by pressing it,
(H) A state display area A08 for displaying the operating state of the control device 10, including displaying visual alarm information under certain conditions.
(I) Criteria for displaying the most recent value of _{rFiO 2 FiO 2} display area A09,
(J) Eupoxya time display area A10, which displays the ratio of time in eupoxia,
_{(K) FiO 2} feedback display area A11, for example, an optical indicator _{indicating whether the servo FiO 2} or the measured FiO ₂ matches the output FiO 2 value.
_{(L) Effective SpO 2} display area A12, for example, an optical indicator indicating whether or not _{the input SpO 2 value is valid.}
(M) Additional input display areas A13 and A14, for example, indicating whether or not signals representing respiratory velocity and respiratory circuit pressure have been input into the control device 10 by optical indicators.

A general control process of a method for automatically controlling inhaled oxygen distribution performed by a controller 11 according to an embodiment is shown in the flowchart of FIG.

As shown in FIG. 12, when the control process is started, the input to the control device 10 is processed in step S1202.

FIG. 13 shows an exemplary step S1300 that processes the input in step S1202. The input is read from the input unit (step S1302) and then validated (step S1304). After validation of the input, a manual override evaluation is performed (step S1306).

An exemplary step of reading the input in step S1302 is shown in FIG. As shown in FIG. 14, in step S1402, an input signal including signals representing _{a plurality of input oxygen saturation (SpO 2) values for the patient is received.} Input signals include SpO ₂ , HR _press , respiratory velocity, CPAP circuit pressure, servo FiO ₂ , measurement FiO ₂ , maximum ΔFiO ₂ , target SpO ₂ , and target range approach target.

An exemplary step of verifying the input in step S1304 is shown in FIG. First, the validation of the input boundary is performed in step S1502, after which the perfusion index state is updated in step S1504. Then, _{the hierarchical verification of SpO 2} is executed in step S1506. Also, in step S1508, an alarm is triggered based on the result of Hierarchical verification of _{SpO 2 and whether the servo FiO 2} or measured FiO _{2 deviates from the FiO 2} set beyond the permissible limit. ..

The following steps are used to activate the alarm in step S1508.

(A) If the SpO ₂ signal is level 1 or level 3 for more than 30 seconds, the missing signal = true alarm is activated.
If the SpO ₂ signal is level 1 or level 3 for more than 2 minutes, increase the alarm volume of the missing signal and display an error message to check the oximeter probe and connection.

(B) If | (Setting FiO ₂ -Servo FiO ₂ ) / (Setting FiO ₂ ) |> 1%, the alarm is activated assuming _{that Servo FiO 2 mismatch = true.}
If | (setting FiO ₂ -servo FiO ₂ ) / (setting FiO ₂ ) |> 5%, the _{alarm is activated assuming that the servo FiO 2} error is true and the manual mode is true. Let me.

(C) If | (Setting FiO ₂ -Measurement FiO ₂ ) / (Setting FiO ₂ ) |> 2%, the alarm is activated assuming _{that measurement FiO 2 mismatch = true.}
If | (Setting FiO ₂ -Measurement FiO ₂ ) / (Setting FiO ₂ ) |> 10%, the _{alarm is triggered assuming that measurement FiO 2} error = true and manual mode = true.

Further, FIG. 16 shows an exemplary step of verifying the input boundary in step S1502.

As shown in FIG. 16, first, the effective input boundary is set in step S1602. For example, the valid input boundaries for IB (1) -IB (9) are set as follows.
IB (1) SpO ₂ : 0 ≤ SpO ₂ ≤ 100%
IB (2) HR _pleth : 0 ≤ HR _pleth ≤ 300 bpm
IB (3) Perfusion index: 0 ≤ perfusion index ≤ 10
IB (4) Pretismograph waveform: 0 ≤ plethysmograph waveform ≤ 5V
IB (5) HR _ecg : 0 ≤ HR _ecg ≤ 300 bpm
IB (6) Respiration rate: 0 ≤ Respiration rate ≤ 150 / min IB (7) CPAP circuit pressure: 0 ≤ CPAP circuit pressure ≤ 20 cmH ₂ O
IB (8) Servo FiO ₂ : 21% ≤ Servo FiO ₂ ≤ 100%
IB (9) Measured FiO ₂ : 21% ≤ Measured FiO ₂ ≤ 100%

Next, through the loop of steps S1604, S1608, and S1616, it is determined whether or not the input value of each IB (i) is outside the valid input value boundary. If not, the loop proceeds to determine the next IB (i). If so, the logic moves to step S1610, where the most recent valid value of IB (i) is used in place of the current IB (i). After step S1610, a test for determining whether an extension timeout has occurred is performed in step S1612. If yes, in step S1614 the input timeout flag ITO (i) is set and the logic proceeds to step S1616 to process the next IB (i) or end the input boundary evaluation.

When all IBs (i) have been verified, the logic moves to step S1618 to determine if the alarm should be triggered based on the current value of CPAP circuit pressure, for example using the following steps: ..

a. Verification that the circuit pressure is within range If the CPAP circuit pressure is less than 50% of the plateau value, the CPAP circuit pressure = low (ie, if the circuit pressure is less than 50% of the plateau value, the circuit Judge that the pressure is low)
If the CPAP circuit pressure is _{less than 1 cmH 2} O, the CPAP circuit pressure = loss (ie, if the circuit pressure is _{lower than the plateau value of 1 cmH 2} O, the circuit pressure is determined to be completely lost).

b. Activating the CPAP circuit pressure alarm If the CPAP circuit pressure is lower than 30 seconds, the CPAP circuit pressure drop alarm is activated. (That is, if the circuit pressure is 50% or less for 30 seconds or more, an alarm is triggered.)

In step S1504, updating the perfusion index state employs the process shown in FIG. 6 described above.

In step S1506, _{the hierarchical verification of SpO 2} employs the process shown in FIG. 7 described above.

In step S1306, the manual override evaluation employs the process shown in FIG. 9 described above.

After processing the input in step S1202, automatic control is executed in step S1204 to determine _{the FiO 2} value based on the _{input SpO 2 value.}

FIG. 17 shows an exemplary step of automatic control in step S1204. As shown in FIG. 17, when the automatic control is started, the periodic addition process is executed first in step S1702.

An exemplary step of the periodic addition process in step S1702 is shown in the flowchart of FIG. As shown in FIG. 18, the periodic addition process includes _{updating the reference FiO 2 in} step S1802 and updating the performance analysis in step S1804. They employ the exemplary processes shown in FIGS. 5 and 4, respectively.

After the periodic addition process in step S1702, the PID term is generated in step S1704. An exemplary process as shown in FIG. 3 described above may be employed in step S1704.

In step S1706, the output FiO ₂ value is determined based on the validity of the PID term, rFiO ₂ value and input SpO _{2 value.} An exemplary process, as shown in FIG. 8 above, is employed to determine the _{output FiO 2 value in step S1706.}

In step S1708, it is detected whether the device has been switched to manual mode. If it is detected that the device has been switched to manual mode, the output FiO ₂ value is set to be equal _{to the FiO 2} value selected by the user (bedside staff).

In step S1710, the control process determines if one or more alarms need to be triggered and controls the alarms accordingly. The exemplary process shown in FIG. 10 described above is adopted by step S1710.

In FIG. 12, after step S1204, the determined FiO ₂ value is set as an output and transmitted to the output unit in step S1205. For example, the control process _{displays data including the received input SpO 2} and the updated output, updates the display, and reflects the updated data (step S1208).

After updating the display, the control process detects whether or not an existing user input instructing automatic control has been detected (step S1210). If not detected, the control process proceeds to step 1202 and repeats steps S1202 through S1210 again. If the user input instructing the existing automatic control is detected, the control process ends.

Further, in a certain embodiment described above, a method for automatically controlling inhaled oxygen distribution is executed by the control device 10, but the method includes software, hardware, firmware, software, hardware, and. / Or a combination of firmware and / or may be performed in the form of other mechanisms. For example, the method may be performed, for example, in machine-readable code generated using a coding tool, by the steps of running a computer or microcomputer. The software may be integrated with or installed in a controller, oximeter, or respiratory support device. The signal described here is an electronic signal, and the stored value is stored in an electronically accessible non-volatile storage.

Described here is a device for automatically controlling inhaled oxygen distribution.

The apparatus includes an input unit for receiving a signal representing a plurality of input oxygen saturation (SpO ₂₎ values for patients, a memory for recording the input SpO ₂ values received, based on the input SpO ₂ values output inspired oxygen (FiO ₂ ) It has a controller that determines the value and an output unit that outputs the _{determined output FiO 2 value.}

The controller generates a control value based on the input SpO ₂ value and the target SpO ₂ value, and generates an output FiO ₂ value based on the control value and the reference inspired oxygen concentration (rFiO _{2) value.} As described above, the control values are based on the input SpO ₂ value, the target SpO ₂ value, and the immediate control value, the input SpO ₂ value, the target SpO ₂ value, and the cumulative gain coefficient generated based on the immediate gain coefficient. The immediate gain coefficient is determined based on _{the rFiO 2} value and includes the cumulative control value, the input SpO ₂ value, the target SpO ₂ value, and the predicted gain coefficient generated based on the predicted gain coefficient. Non-linear weighted compensation is applied to the cumulative control value based on a predetermined non-linear relationship between the partial pressure (PaO ₂ ) and SpO _2.

For example, the device has the configuration of the control device 10 as shown in FIG.

The system described here is a system for automatically controlling inhaled oxygen distribution. The system communicates between one or more oxygen saturation monitoring devices and one or more inhaled oxygen control devices, control devices, and one or more oxygen saturation monitoring devices and control devices, and one or more. It has a network that allows communication between the inhaled oxygen control device and the control device.

_{The control device receives signals representing multiple input oxygen saturation (SpO 2} ) values for the patient from each of the one or more oxygen saturation monitoring devices over the network and controls based on the _{input SpO 2} value and the target SpO _{2 value.} Generate values, generate output inspiratory oxygen saturation (FiO ₂ ) values based on controlled values and reference inspired oxygen concentration (rFiO ₂ ) values, and transfer the determined output FiO ₂ values to the corresponding inhaled oxygen control device over the network. Control oxygen distribution by transmitting. As described above, the control values are based on the input SpO ₂ values, the target SpO ₂ values, and the immediate control values generated based on the immediate gain coefficients, the input SpO ₂ values, the target SpO ₂ values, and the cumulative gain coefficient. The immediate gain coefficient is determined based on the _{rFiO 2} value, including the cumulative control value generated in the above and the input SpO ₂ value, the target SpO ₂ value, and the predicted gain coefficient generated based on the predicted gain coefficient. Non-linear weighted compensation is applied to the cumulative control value based on a predetermined non-linear relationship between the partial pressure of arterial oxygen (PaO ₂ ) and SpO _2.

In this way, the control device is used within the network with one or more pairs of oxygen saturation monitoring devices and inhaled oxygen control devices connected to the network. This enables real-time automatic control of performance at remote sites, enabling large-scale data collection. The central controller also simplifies the adjustment or modification of the control process.

Described below are exemplary experiments and corresponding experimental results related to methods of automatically controlling inhaled oxygen distribution.

<First Example>
(Method)
In the first example, a proportional integral differential (PID) controller _{compensates for (i) a non-linear SpO 2-} PaO ₂ relationship, (ii) adapts to the severity of pulmonary dysfunction, and (iii) makes an error within the target range. Is strengthened by attenuating.

The oxygen control method was performed on a stand-alone device consisting of a processing platform (laptop computer), device inputs and outputs, a servo-controlled air-oxygen mixer, and a user interface displayed on a computer screen. The control instructions were written in a graphical programming language (LabVIEW2010, National Instruments, USA, by Austin) and uploaded into a laptop computer as machine-readable instructions.

The instructions provided a proportional integral differential (PID) controller. For PID control, an error is defined as a deviation of the processing signal from the setpoint. The value of the manipulated signal output at each moment is proportional to the error, its integral, and its derivative, and has different multiplier coefficients in each case (K _p , _Ki , K _d ). In this case, the error (e) is _{the numerical difference between the input value of SpO 2} (assuming a valid signal) and the midpoint of the selected target range (for example, the target range 91-95% and the midpoint 93%). Met. The integrand (∫edτ) is the sum of all errors (subject to the limitations shown below), and the integral term in PID control helps to overcome certain state errors. The derivative (de / dt) is the SpO ₂ gradient by linear regression over the preceding 5 seconds, and PID control gives a prediction of future errors. At each iteration, the output of the process _{was ΔFiO 2} , which is the sum of each of the PID terms (Equation 4). _{The FiO 2} to be distributed (set FiO ₂ ) is the sum of ΔFiO ₂ and the reference FiO ₂ value (rFiO ₂ ) and represents the current baseline oxygen requirement (Equation 5). Setting FiO2 ₂ , rounded to ± 0.5% and forced to a value between 21 and 100%.

The PiD control process is in a loop that repeats every second, and FiO ₂ changes are made in 1 second if necessary. The range of values for K _p , _Ki and K _d was derived from extensive simulation studies. The value of _K p, _{K i} and _{K d} used in this example, _{K p} is -1, _{K i} is -0,0125, _{K d} was -1. Values of K _p were adapted in the range between -0.5 and -1 depending on the severity of pulmonary dysfunction (see below).

Modifications to the PID controller are adapted to accommodate certain idiosyncratic constitutions of the controlled system. _{The errors associated with SpO 2} values within the target range were reduced by applying a multiplier proportional to the distance from the midpoint of the target range (target range attenuation). _{Also, given a relative inaccuracy of SpO 2} monitoring with a value of 80% or less, negative errors were capped at 13%. These error adjustments were applied only to the calculation of proportional terms.

A modification has been made to the handling of integral terms. _{In the recognition that the integral term gradually increments FiO 2} in the event of long-term hypoxia, its magnitude was capped to limit _{the maximum ΔFiO 2} to 40% above _{rFiO 2. In hyperoxygenosis (SpO 2} beyond the target range when oxygenated), it follows hypoxia as an overshoot, and _{errors at high SpO 2} values are not proportional to similar deviations _{of PaO 2} from tolerance. (Fig. 19).

To overcome this, severing house compensation is adopted, thereby multiplying the positive error entered by the error multiplier as long as the _{integral term is positive (ie, the tendency to increase ΔFiO 2) during excess oxygen.} Was applied (see Table 2 below). In determining the integral term, the error multiplier was applied _{to the positive SpO 2} error until the integral term was reduced to zero. The value of the error multiplier was derived from the Seberinghouse equation. In the room air, the integral term was not changed when continuous values of _{SpO 2 above the target range were no longer considered to represent constant hyperoxia.}

The differential term calculation is also modified in hyperoxygenosis, resulting in _{a negative SpO 2} gradient being null (ie equal to 0) _{if all five recent SpO 2} values exceed the setpoint. Was done. Therefore, the pressure above _{ΔFiO 2} due to the differential term was avoided in hyperoxygenosis.

K _p is modified in accordance with the seriousness of pulmonary dysfunction by applying a scaling factor proportional to RFIO ₂ current, additional approaches have been investigated. The K _p modification was performed by multiplying the current value of _{K p by} a factor in the range of 0.5 to 1.0 for _{rFiO 2} in the corresponding range of 21% to 40%. The application of such K _p recognizes an inverse proportional relationship between the severity and gain of pulmonary dysfunction observed at this rate.

The primary input SpO ₂ to the control process is sourced from any oximeter having an analog or digital output. For preclinical testing, SpO ₂ was derived from a simulation of oxygen supply in preterm infants. The output from the control process is transmitted to any device that receives and executes the _{desired value of FiO 2} , including an air oxygen mixer and a ventilator. For preclinical testing, output FiO ₂ was linked to an oxygen supply simulator.

(Preclinical test)
The contributions of the three enhancement features were investigated. a) SEBA-House compensation, b) K _p applied, and, c) all permutations performance PID control with the target range attenuation, was evaluated using a simulation of the oxygen supply. _{1 Hz recordings (24 hours) of FiO 2} and SpO ₂ from each of the 16 preterm infants at sustained positive airway pressure were converted to continuous values for ventilation / perfusion ratio and shunt. The SpO ₂ average time of the original recording was 2-4 seconds and was not averaged during the data abstraction and simulation. The V / Q and shunt sequences were linked to the controller under tests within an automated oxygen controller to generate a series of unique values for _{SpO 2.} The SpO ₂ target range was set to 91-95%. Non-integral (ie, proportional derivative, PD) controller functionality and fully enhanced controller functionality with a 30 second lockout after _{FiO 2 adjustment were also tested.} For the latter analysis, multiple permutations of PID coefficients were tried to optimize performance.

For each of the 16 SpO ₂ sequences generated during the simulation, the percentage of time in the following oxygen supply states was calculated. That is, SpO _{2 of the target range, eupoxia (SpO 2} exceeding the target range or the target range in the case of indoor air), SpO ₂ <80%, SpO ₂ <85%, above or below the target range, 96%. In excess of oxygen, in excess of 98% oxygen. Prolonged hypoxia _(SpO 2 <85%) and hyperoxia identifies the frequency of expression of (oxygen SpO ₂ greater than 96%), similar to the frequency of SpO ₂ overshoot, _SpO 2 <85% of It is defined as a _{SpO 2} reading over the target range for at least 60 seconds over the 2 minutes following the hypoxia event. SpO ₂ instability was _{assessed using changes in SpO 2} coefficient (CV), frequency, and mean interval of expression outside the target range. These data are _{summarized as median and interquartile range (IQR) for non-SpO 2} overshoots, where the data are pooled and represented as a single value for each control process. Control performance was assessed by a median comparison using Friedman's nonparametric iterative measurements (ANOVA and Dan's post-hook test). For simplicity, the comparison was limited to the following set: a) either without PID having a single reinforcement factor (SEBA-House compensation / _{K p} adaptation / target range attenuation), b) one not having reinforcing PID or having a subtraction strengthening factor, c) 30 sec lockout / Comparison of PD and PID / enhanced PID / PID. _{Summary data related to manually controlled SpO 2} goals from the original record was also generated, but _{no statistical comparisons were made for the different SpO 2} target ranges given (88-92%).

(result)
The records used in the simulation were born during a 30.5 week gestation period (IQR 27.5-31 weeks), had a birth weight of 1320 (910-1860) grams, and 16 on 2 (0-5.3) days after birth. It was taken from a human preterm baby. Infants have a significant degree of SpO _{2 instability and have symptoms of hypoxia (SpO 2} <80) with a frequency of 3.1 (1.6-9.9) occurrences every 4 hours. It was expressed. At the time of recording, the CPAP pressure level was 7.0 (6.5 to 8.0) cmH ₂ O and the baseline FiO ₂ range was 0.21 to 0.61. After removal of the missing SpO ₂ data, the recording was between 22 (20-26) hours.

In the simulation test, the complementary functions of the different components of the PID controller were clear. Separate addition of K _p application and target range attenuation to the PID controller has been improved Eupokishia time, whereas sebacate-House compensation reduced the expression of hyperoxia (see Tables 3 and 4). Overall, the performance of the PID controller with all three enhancements was superior to the other combinations. In the absence of target range attenuation, epoxia time tended to be longer than for fully enhanced PID (see Table 3). The development of hypoxia and hyperoxygenosis was _{most effectively eliminated without Kp} indication (see Table 4). Removal of sevaling house compensation from the enhanced controller minimized hypoxia, but led to more time and more onset of hyperoxygenosis cyclically (see Tables 3 and 4). .. Enhanced controllers were, in all respects, _{better than controllers with a lockout period of 30 seconds after each FiO 2} change and significantly better than PD controllers (see Tables 3 and 4).

The stability of the SpO ₂ record also varied significantly with different permutations of enhancement features (see Table 5). The SpO ₂ CV values being recorded generally reflected the instability found in the individual examples (eg, with the _Kp adaptation removed from the enhanced controller). SpO ₂ CV is minimized by the enhanced controller (and some other combinations), suggesting that it is relatively stable under these circumstances. Both PID and PD controls with a 30 second lockout _{reduced SpO 2} stability with longer onset above and below the target range (see Table 5).

The separate addition of each enhancement feature to the PID controller has shown benefits. Reinforcement controllers performed better overall than PID controllers with less enhancement, with optimal time combinations and avoidance of hypoxia and hyperoxia in _{the desired SpO 2 range.} This controller performed better than the one with a 30 second lockout and significantly better than the PD control.

The enhanced PID controller was able to respond quickly to _{SpO 2} deviations, adjusting _{FiO 2 up} to once per second if necessary. The initial response to hypoxia or hyperoxygenation events was largely in the domain of proportional and differential terms, with _{more relaxed FiO 2 regulation determined by integral terms until oxygen normal.}

At least in simulation, enhanced PID controllers were very effective in alleviating the development of long-term hypoxia and hyperoxygenosis. The addition of severing house compensation to the PID controller helped to overcome hyperoxygenosis events (including overshoots). Removing it from the reinforced controller caused its recurrence.

In summary, in the preclinical test using oxygen supply simulation, the enhanced controller was _{very effective in targeting the desired SpO 2} range and avoiding oversupply of oxygen.

予め特定されたSpO₂範囲内の時間割合(全時間の%)の比較、メジアン (四分位範囲)、カラム内統計比較 (フリードマン ANOVA with Dunn’s ポストホックテスト)、 ａ：ｂからの差, P<0.05、 ｃ：ｄからの差、 ｅ：ｆからの差、 ｇ：ｈからの差、ｉ：ｊからの差、ｋ：ｌからの差、 PID: 比例積分微分、K_p: 比例係数、 SC: セバリングハウス補償、 TRA:目標範囲減衰

Comparison of time percentages (% of total time) within a pre-specified SpO ₂ range, median (interquartile range), intra-column statistical comparison (Friedman ANOVA with Dunn's post-hook test), difference from a: b, P <0.05, c: difference from d, e: difference from f, g: difference from h, i: difference from j, k: difference from l, PID: proportional integral derivative, K _p : proportional coefficient, SC: Severing house compensation, TRA: Target range attenuation

低酸素症および酸素過剰症発現（＞３０秒および＞６０秒間隔）の連続頻度およびオーバーシュートの頻度の比較、カラム内統計比較 (フリードマン ANOVA with Dunn’s ポストホックテスト)、 ａ：ｂからの差、 P<0.05, ｃ：ｄからの差、 ｅ：ｆからの差、 ｇ：ｈからの差、＊：１６個のすべての記録に対してプールされたオーバーシュート発現のデータ. 表３に従う省略、オーバーシュートの定義に対する上述した方法を参照

Comparison of continuous and overshoot frequencies of hypoxia and hyperoxia (> 30 and> 60 second intervals), intracolumn statistical comparison (Friedman ANOVA with Dunn's post-hook test), difference from a: b, P <0.05, c: difference from d, e: difference from f, g: difference from h, *: data of overshoot expression pooled for all 16 records. Omitted according to Table 3. See the method described above for the definition of overshoot

SpO₂ 不安定性の指標. メジアン (四分位範囲)、カラム内統計比較 (フリードマン ANOVA with Dunn’s ポストホックテスト)、 ａ：ｂからの差、 P<0.05、 ｃ：ｄからの差、 ｅ：ｆからの差、 ｇ：ｈからの差、 ｉ：ｊからの差、 PID: 比例積分微分、 K_p: 比例係数、 SC: セバリングハウス補償、TRA:目標範囲減衰

Indicator of SpO ₂ instability. Median (interquartile range), intra-column statistical comparison (Friedman ANOVA with Dunn's post-hook test), a: difference from b, P <0.05, c: difference from d, e: f Difference from, g: Difference from h, i: Difference from j, PID: Proportional integral derivative, K _p : Proportional coefficient, SC: Interquartile range compensation, TRA: Target range attenuation

<Second Example>
(Method)
In a second example, the enhanced PID controller of Example 1 was incorporated into an oxygen control device and tested by clinical evaluation.

As shown in FIG. 20, devices incorporating automatic oxygen control methods include laptop computers, automatic air-oxygen mixers, and data input / output devices incorporating analog-to-digital (AD) converters (USB-6008, National Instruments, Austin, It was a stand-alone device consisting of USA). The controller is a standard cardiopulmonary monitor (Drager Infinity, Drager Medical Systems Inc, Notting Hill, Australia), SpO ₂ (Masimo Corp, Irvine, California), and heart rate (HR) determined from the electrocardiogram signal. _ecg ) and digital inputs including plethysmograph heart rate (HR _pleth ) were received. The SpO ₂ average was set immediately (2-4 seconds). FiO ₂ was measured via a sensor in the proximal rim of the respiratory circuit (Teledyne) and input to the device via an AD converter. _{The desired value of FiO 2} derived from the controller is directed to the servomotor (model HS-322HD, Hitec RCD USA, Poway, USA) custom-mounted on the air-oxygen mixer (Bird Ultrablender, Carefusion, Seven Hills, NSW). It was attached. As a result, the FiO ₂ selection dial of the mixer can be automatically rotated through the annular gear mechanism. The servomotor and gear system had sufficient torque and accuracy to allow small adjustments (minimum ± 0.5%) to be made accurately and repeatedly with respect to _{FiO 2.} The servomotor has a low holding torque so that the dial of the mixer can be turned manually. This manual intervention was detected by a position sensor _{and caused a switch to manual mode where FiO 2} was no longer under automatic control (see below). At the beginning of each study, servomotor calibration was checked and changed if necessary.

The automatic control method consists of a PID control process with enhancements in the determination of proportional, integral, and derivative terms to properly apply PID control to the automatic oxygen control of preterm infants. The enhancement of the proportional term includes the severity of pulmonary dysfunction, error attenuation during the target range, and modulation based on the error upper bound during hypoxia. Strengthening of the integral term, the upper limit specified in the size of the integrand, compensation for non-linear relationship PaO ₂ -SpO _2, and includes a restriction of the integrand increase in room air.

The PID control process repeated a loop of up to once every second. Therefore, the method was designed to detect and respond to very frequent, rapid changes in oxygen supply in preterm infants. The range of PID coefficient values was derived from extensive simulation studies using data from preterm infants so that multiple permutations of different values were tested for all coefficients. The values of _{K p} , K _i , and K _d used in the examples _{were K p} -1, K _i −0.0125, and K _d -1. Values of K _p were adapted in the range between -0.5 and -1 depending on the severity of pulmonary dysfunction.

Non-numeric SpO ₂ value is treated as missing, in which case _the value of _{HR ecg} and _{HR pleth} were SpO ₂ different places beyond 30 bpm. In the Missing SpO ₂ value event, FiO ₂ was retained at the current value. Full functionality of the controller was resumed immediately after the active signal was restored.

During automatic control, bedside staff could override the control device by manually rotating _{the FiO 2 dial on the mixer.} It signals manual override through detection of a discrepancy between the setting FiO _{2 and the FiO 2} value detected by the position sensor in the servomotor. Becomes a manual override, the 30 seconds after the last manual changes to the FiO _2, the automatic control is resumed with FiO ₂ selected by the user. The device could be locked into manual control mode, if necessary, by a research term of instructions from bedside staff.

(Clinical trial)
The study was conducted in the neonatal and pediatric intensive care unit at Royal Hobart Hospital. The device treats approximately 70 preterm infants with a gestation period of less than 32 weeks each year, and this group of patients is treated with continuous positive airway pressure (CPAP) and high-flow nasal canula (HFNC) whenever possible. ) Has the spirit of using non-invasive respiratory support. _{The SpO 2} target range for oxygen therapy titrations has been updated from the previous 88-92% to 90-94%.

Pregnant infants with a gestation period of less than 37 weeks and less than 4 months of age are suitable for study if they receive non-invasive respiratory support (CPAP or HFNC) at the beginning of the study period and receive supplemental oxygen. rice field. Infants with significant instability or congenital anomalies (including heart malformations other than patent ductus arteriosus) were excluded.

This was a 4-hour predictive intervention study of automatic oxygen control. It was compared to two adjacent periods of standard manual control for a total of 8 hours (4 hours before and after automatic control). There was a 15 minute interval between study periods to avoid carry-over effects. The researchers were present during the automatic control period, but did not interfere with bedside medical staff unless the clinical system malfunctioned. During the period of automatic control, the caregiver was able to override the control device output with the customized air-oxygen mixer used in the study by turning the dial on the mixer. During manual control recording, bedside caregivers were instructed to use the usual approach to _{SpO 2 goals with a standard SpO 2} goal range (90-94%). Based on previous studies, manual control prioritized the upper limit of this range. Given that automatic control targets the midpoint of the SpO ₂ range, the target range is set to 91-95% during automatic control, and manual and automatic SpO ₂ histograms are expected to overlap. For both manual and automated study periods, the alarm settings were the same with a lower limit of 89% and an upper limit of 96%.

Prior studies oximetry probe is disposed at a position after the conduit, lowering of medical need or continuous SpO ₂ signal unless occur not moved between the three study periods. Care times were scheduled to be outside the data recording time, if possible. For automatic control, _{a constant value of reference FiO 2} (rFiO ₂ ) was selected for each infant based on the most recent basic supplemental oxygen requirement.

Relevant demographic and clinical data were recorded for each infant, including gestational age, birth weight, level of respiratory support at the time of study, and details of clinical status. SpO ₂ and FiO ₂ were recorded at 1 Hz during both manual and automatic control. Analysis of these records made it possible to assess _{SpO 2 instability in each infant by assessing changes in the SpO 2} coefficient, the number of expressions outside the target range, and the mean time interval. In addition, the ratio of each time of the following oxygen supply states was confirmed. SpO ₂ target range, Eupokishia (the target range SpO _2, or, _SpO 2 above the target range in the room air), SpO ₂ alarm range (89-96%), and, less than 80% SpO _2, oxygen Is over 80-84%, 85-88%, 97-98%, and 98%. To calculate these values, the denominator was available for the time after data exclusion during the time interval _{of the missing SpO 2 signal.} The frequency of long-term onset of hypoxia and hyperoxia is identified as the frequency of _{SpO 2} overhead and is above the target range for at least 60 seconds over the 2 minutes following the event of hypoxia with _{SpO 2 <85%.} Was defined as a SpO ₂ read of. _{The number of FiO 2} adjustments between manual and automatic recording was determined as the mean oxygen exposure (mean FiO _{2) in each case.}

Data were expressed as median and interquartile range (IQR), unless otherwise stated. The comparison was performed between the automatic control time and the manual control time using the Wilcoxon match pair test. For these analyzes, analytical data from manual control time is pooled, but in addition, for each infant, 2 hours or more manual with optimal manual control time (ie, the maximum percentage of time in eupoxia). Recording interval) was also used as a comparator. The primary result was the percentage of time in eupoxia. The sample size selected for the study was primarily based on the need to increase the initial medical experience of controller performance and increase the safety of a sufficient number of subjects. In a previous study of 45 infants, we found a percentage of time within the target range if oxygen should be 30 ± 15% (mean ± SD). Assuming a similar standard deviation for the difference between the paired automatic and manual control values in this study, a sample of 20 infants produced 80% power and 0.05 α error. It has become possible to detect a difference of 10% in the eupoxia time between the manual time and the automatic time.

(result)
The study was conducted from May to December 2015. The participating infants (n = 20) had a gestation period of 27.5 weeks (IQR 26-30 weeks) and a birth weight of 1130 (940-1400) grams. Of the 20 infants, 15 (75%) were boys. Infants were studied at 8.0 (1.8-3.4) days after birth, with a corrected gestational age of 31 (29-33) post-menstrual weeks and a baby weighing 1400 (1120-1960) grams. there were. For infants (n = 13) studied for CPAP, the pressure level at the start of recording was 6 (5-8) cmH ₂ O. For infants (n = 7) studied for HFNC, the starting flow rate was 6 (5.5-6.5) L / min. In all cases, the ratio of nurses to patients was 1: 2.

Data from two adjacent periods of manual control are available from 18 infants, the need for intubation immediately after automatic control and the use of a second manual control data log (on a case-by-case basis) for data logging failures. That was the reason why it was impossible. The percentage of missing signals is 2.9 (0.5 to 5.4)% and 1.7 (0.7 to 3.4)% in each of the first manual, automatic, and second manual recordings. And 1.5 (0.8 to 7.1)%, and the time available for analysis is 3.8 (3.7 to 4.0) and 3.8 (3.7 to 3.9). ) And 4.0 (3.8-4.0) hours left.

21A and 21B show 2-hour recordings from the same infant under manual and automatic control. It includes _{sample recordings of SpO 2} (solid line, Y-axis:% saturation) and FiO ₂ (dashed line, Y-axis:% oxygen) below.
(A) Infant 5 was born in a gestation period of 27 weeks and was studied 40 days after birth, with a high flow nasal canula (6 L / min), manually controlled, eupoxia time 59% (see Figure 21A).
(B) Infant 5 has an eupoxia time of 79% by automatic control and an eupoxia time of 82% in a state of being automatically controlled by _{rFiO 2 set to 29% throughout (see FIG. 21B).}

21A and 21B reveal typical variability of _{SpO 2} during manual control (FIG. 21A) and automatic control of _{FiO 2 (FIG. 21B).} It was less noticeable during automatic control (FIG. 21B). The exemplary data shown in FIGS. 21A and 21B show _{the rapid response of FiO 2} produced by the controller and the increase in time in the target range (shown by the gray band).

In FIG. 22, the black bar indicates manual control, the white bar indicates automatic control, T indicates SpO ₂ within the target range, the target range is 90 to 94% for manual control, and 91 for automatic control. ~ 95%. The frequency histogram of the pooled SpO ₂ data shows a substantial increase in the percentage of time within the target range due to automatic control, and both hypoxia and hyperoxygenosis values are shown below compared to manual control. ing. _{The SpO 2} target profile under manual control appears to have a curved peak at the top of the target range. In contrast, as expected, automatic control targeted the midpoint of the selected target range (ie, SpO ₂ : 93%). When supplemented oxygen was received, median SpO ₂ in the pooled data was 93% for both manual and automatic control.

_{Oxygen supply has a slight SpO 2} deviation below the target range and below 80% during automatic control, much more with all manifestations outside the target range at shorter time intervals compared to manual control. It was stable. Fluctuations in the SpO ₂ coefficient were also quite different (3.8 (3.2-4.7)% for manual control, 2.3 (1.8-3.0%, P <0.0001) for automatic control. )).

Compared to manual control time, automatic control resulted in 23% and 25% more time in the target and eupoxia ranges, respectively (see Table 6). I also spent more time within the alarm range (89-96%). Automatic control reduced significant time at both poles of oxygen supply and substantially eliminated SpO ₂ <80% hypoxia and SpO ₂ > 98% hyperoxygenosis. The shorter range of time required for hypoxia and hyperoxygenosis was also reduced.

These findings were reflected in the analysis of the long-term manifestations of hypoxia and iatrogenic hyperoxia. Both occur reasonably frequently during manual control (see Table 7), but were clearly rare during automatic control. No overshoot was identified in any of the automated control records.

23, the value of the time of personal Eupokishia for optimal manual control time is compared with the automatic control, horizontal rod is median, Eupokishia exceeds SpO _2, or the target range room air target range SpO ₂ Is. There is a clear benefit of automatic control when measured against the optimal two manual control times, with 60 (50-72)% of the time in the eupoxia range and 60 (50-72)%, respectively, for optimal manual control and automatic control. It was 81 (76-90)% (P <0.0001). In addition, automatic control was _{associated with better SpO 2} goals in individual studies, with relative improvement over the eupoxia time range of 2.2 to 55% (see Figure 23).

During the manual control time, at least 1% FiO ₂ adjustment was performed 2.3 (1.3 to 3.4) hourly by bedside staff. During automatic control, a 0.5% FiO ₂ minimum change is periodically (overall 9.9 changes per minute) actuated by a servomotor, measuring at least 1% at a frequency of 64 (49-98) per hour. A change to FiO _{2 has occurred.} A total of 18 manual adjustments were performed (0.24 adjustments per hour) in all 20 records of automatic control re-regulation, and a 90% reduction from the ratio of manual adjustments was observed during manual control. (2.3 times per hour). The maximum number of manual adjustments for individual patients during automatic control was 4 times in a 4-hour record (ie, once per hour). No significant system dysfunction occurred.

Median values of oxygen requirement (mean FiO ₂ ) were 27 (25-30)%, 27 (25-30)%, and 26 (24) relative to the first manual, automatic, and second manual recordings, respectively. It was ~ 31)%. Oxygen requirements were not different between automatic and either manual recording (P> 0.05, Wilcoxon match pair test).

In short, the enhanced PID controller was more effective than routine manual control by 25% or more within the desired SpO _{2 range in the SpO 2 target.} Extreme oxygen supply was largely avoided and long-term hypoxia and hyperoxygenosis were virtually eliminated. Effective oxygen control was achieved with similar exposure to oxygen, with little manual fraction of inhaled oxygen regulation.

手動および自動制御に対して、予め特定されたSpO₂範囲内の時間割合（全利用可能時間の％）の比較、２つの隣接する期間からプールされた手動制御データ、メジアン (四分位範囲)、＊：ウィルコクスンマッチペアテスト

_{Comparison of time percentages (% of total available time) within a pre-specified SpO 2} range for manual and automatic control, manual control data pooled from two adjacent periods, median (interquartile range) , *: Wilcoxon Match Pair Test

手動および自動制御間の低酸素症および酸素過剰症発現（＞３０秒および＞６０秒間隔）の連続頻度の比較、２つの隣接する期間からプールされた手動制御データ、メジアン (四分位範囲)、＊：ウィルコクスンマッチペアテスト

Comparison of continuous frequency of hypoxia and hyperoxygenosis (> 30 s and> 60 s intervals) between manual and automatic control, manual control data pooled from two adjacent periods, median (interquartile range) , *: Wilcoxon Match Pair Test

As described with reference to the accompanying drawings, it is known to those skilled in the art that many modifications can be made without departing from the aspects of the present invention.

Related Application This application claims the priority of Australian Provisional Patent Application No. 2015904621 filed on November 10, 2015, which is incorporated herein by reference.

Claims (18)

A method of operating a device for automatically controlling inhaled oxygen distribution , wherein the device has an input unit, a controller, and an output unit.
A step in which the input unit receives signals representing a plurality of input oxygen saturation values (input SpO _{2 values) having different times for a patient, and}
A step in which the controller generates a control value based on the input SpO ₂ value and the target SpO _{2 value, and}
A step in which the controller generates an output inhaled oxygen concentration value (output FiO ₂ value) output from the output unit based on the control value and the reference inhaled oxygen concentration value (rFiO _{2 value).}
With
The control value is
An immediate control value generated by the controller based on the input SpO ₂ value, the target SpO ₂ value, and the immediate gain coefficient.
A cumulative control value generated by the controller based on the input SpO ₂ value, the target SpO ₂ value, and the cumulative gain coefficient.
A predictive control value generated by the controller based on the input SpO ₂ value, the target SpO ₂ value, and the predictive gain coefficient.
Have,
The controller determines the immediate control value based further on the rFiO _{2 value.}
The controller, method based on a predetermined non-linear relationship between the arterial oxygen partial pressure (PaO ₂₎ and SpO _2, applies a nonlinear weighting compensating the accumulated control value, characterized in that. The immediate control value is generated by multiplying the error value by the immediate gain coefficient, and the error value is related to the difference between the _{input SpO 2} value and the target SpO _{2 value.}
The cumulative control value is generated by multiplying the sum or integral of the error values by the cumulative gain coefficient.
The method according to claim 1, wherein the predicted control value is generated by multiplying the difference or the differential coefficient of the error value by the predicted gain coefficient. The target SpO ₂ value is within the target SpO ₂ range selected in advance by the user .
When the current input SpO ₂ value is _{within the target SpO 2} range, the controller applies an attenuator to the immediate control value, which is intermediate between the current input SpO ₂ value and the target SpO _{2 range.} The method of claim 1, characterized in that it is generated by the controller based on the points. When the input SpO ₂ values of the current is lower than the target SpO ₂ value, the error value associated with the difference between the target SpO ₂ value and the input SpO ₂ values of the current, the upper limit of the maximum error value that the user selects The method according to claim 1, wherein the method is defined. The method according to claim 1, further comprising a step of modifying the cumulative control value so as to limit the control value at the maximum control value selected by the user. The method of claim 5, wherein each of the output FiO ₂ values is the sum of the corresponding control value and the corresponding rFiO _{2 value.} The step of controlling the cumulative control value is
(I) Limit the increase in the cumulative control value when the _{current output FiO 2} value is at indoor air level and (ii) the current input SpO ₂ value exceeds the target SpO _{2 value.} The method of claim 1, wherein the method comprises the steps of The method according to claim 1, wherein when the input SpO ₂ value _{exceeds the selected SpO 2 threshold value during the SpO 2} gradient determination period, the predictive control value becomes null. A step in which the controller generates rFiO ₂ evaluation results based on the input SpO ₂ values and the respective output FiO _{2 values over the rFiO 2 evaluation time interval.}
The method according to claim 1, further comprising a step of the controller _{modifying the rFiO 2} value based on the rFiO _{2 evaluation result.} One of a plurality of verify levels in hierarchical verification method, by classifying the input SpO ₂ values of the current, a step of, based on the input SpO ₂ values of the current, the controller generates an SpO ₂ verification result,
The method of claim 1, further comprising a step of the controller determining the output FiO ₂ value based on the SpO _{2 verification result.} The immediate gain coefficient has an initial value, and the initial value is
(I) Selection between -2 and -0.2, or (ii) -1
The method according to claim 1, wherein the method is characterized by the above. The cumulative gain coefficient has an initial value, and the initial value is
(I) Selection between -0.25 and -0.005, or (ii) -0.0125
The method according to claim 1, wherein the method is characterized by the above. The predicted gain coefficient has an initial value, and the initial value is
(I) Selection between -2 and -0.25, or (ii) -1
The method according to claim 1, wherein the method is characterized by the above. The method of claim 1, based on the performance evaluation results, wherein the controller corrects the immediate gain factor, characterized in that. The controller
_{The hypoxia time interval where the input SpO 2} value within the performance analysis time interval is within the range of hypoxia, and the hypoxia time interval, and
A claim characterized in that the performance evaluation result is generated based on at least one of the hyperoxic time intervals where the _{input SpO 2} value within the performance analysis time interval is within the range of hyperoxygenosis. Item 14. The method according to item 14. The target SpO ₂ value is within _{the target SpO 2} range selected in advance by the user.
The controller
The target time interval where the _{input SpO 2} value within the performance analysis time interval is within the target SpO _{2 range, and}
Wherein either performance analysis the input SpO ₂ values within the time interval is within the target SpO ₂ range, or based on the target SpO ₂ range than at the Eupokishia time interval, at least one of the indoor air, the performance The method according to claim 14 , wherein an evaluation result is generated. A device for automatically controlling inhaled oxygen distribution,
An input unit that receives signals representing multiple input oxygen saturation values (input SpO _{2 values) with different times for the patient, and}
A memory for recording the received input SpO _{2 value, and}
A controller that determines the output inhaled oxygen concentration value (output FiO ₂ value) based on the input SpO _{2 value.}
An output unit that outputs the determined output FiO _{2 value,}
With
The controller
A control value is generated based on the input SpO ₂ value and the target SpO ₂ value, and the control value is generated.
_{The output FiO 2} value is generated based on the control value and the reference inspired oxygen concentration value (rFiO _{2 value).}
The control value is
The input SpO ₂ value, the target SpO ₂ value, and the immediate control value generated based on the immediate gain coefficient.
The input SpO ₂ value, the target SpO ₂ value, and the cumulative control value generated based on the cumulative gain coefficient.
It has the input SpO ₂ value, the target SpO ₂ value, and the predicted control value generated based on the predicted gain coefficient.
The immediate control value is further determined based on the _{rFiO 2 value.}
An apparatus characterized in that non-linear weighting compensation is applied to the cumulative control value based on a predetermined non-linear relationship between arterial oxygen partial pressure (PaO ₂ ) and SpO _2. A system for automatically controlling inhaled oxygen distribution,
One or more oxygen saturation monitoring devices and one or more inhaled oxygen control devices,
Control device and
It comprises a network that enables communication between the one or more oxygen saturation monitoring devices and the control device, and communication between the one or more inhaled oxygen control devices and the control device.
The control device is
Through the network, signals representing a _{plurality of input oxygen saturation values (input SpO 2} values) having different times for the patient are received from each of the one or more oxygen saturation monitoring devices.
A control value is generated based on the input SpO ₂ value and the target SpO ₂ value, and the control value is generated.
An output inhaled oxygen concentration value (output FiO ₂ value) is generated based on the control value and the reference inhaled oxygen concentration value (rFiO _{2 value).}
Control the inhaled oxygen distribution by transmitting _{the determined output FiO 2} value to the corresponding inhaled oxygen control device through the network.
The control value is
The input SpO ₂ value, the target SpO ₂ value, and the immediate control value generated based on the immediate gain coefficient.
The input SpO ₂ value, the target SpO ₂ value, and the cumulative control value generated based on the cumulative gain coefficient.
It has the input SpO ₂ value, the target SpO ₂ value, and a predicted control value generated based on the predicted gain coefficient.
The immediate control value is further determined based on the _{rFiO 2 value.}
A system characterized in that non-linear weighting compensation is applied to the cumulative control value based on a predetermined non-linear relationship between arterial oxygen partial pressure (PaO ₂ ) and SpO _2.

Images